Re: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19)
andrewl@xix-w.bengi.exodus.net Mon, 23 December 2002 23:19 UTC
Received: from trapdoor.merit.edu (ietf-mx.ietf.org [132.151.6.1]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id SAA17461 for <idr-archive@ietf.org>; Mon, 23 Dec 2002 18:19:39 -0500 (EST)
Received: by trapdoor.merit.edu (Postfix) id 4799591214; Mon, 23 Dec 2002 18:22:35 -0500 (EST)
Delivered-To: idr-outgoing@trapdoor.merit.edu
Received: by trapdoor.merit.edu (Postfix, from userid 56) id 115F191250; Mon, 23 Dec 2002 18:22:34 -0500 (EST)
Delivered-To: idr@trapdoor.merit.edu
Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 0733491214 for <idr@trapdoor.merit.edu>; Mon, 23 Dec 2002 18:22:33 -0500 (EST)
Received: by segue.merit.edu (Postfix) id D638E5DF2F; Mon, 23 Dec 2002 18:22:33 -0500 (EST)
Delivered-To: idr@merit.edu
Received: from demiurge.exodus.net (demiurge.exodus.net [216.32.171.82]) by segue.merit.edu (Postfix) with ESMTP id 7AC015DE64 for <idr@merit.edu>; Mon, 23 Dec 2002 18:22:33 -0500 (EST)
Received: (from andrewl@localhost) by demiurge.exodus.net (8.9.3+Sun/8.9.3) id PAA24100; Mon, 23 Dec 2002 15:19:24 -0800 (PST)
Date: Mon, 23 Dec 2002 15:19:24 -0800
From: andrewl@xix-w.bengi.exodus.net
To: Tom Petch <nwnetworks@dial.pipex.com>
Cc: idr@merit.edu, andrewl@cw.net
Subject: Re: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19)
Message-ID: <20021223151924.G21226@demiurge.exodus.net>
References: <000e01c2aad8$805f9200$0301a8c0@tom3>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Disposition: inline
User-Agent: Mutt/1.2.5i
In-Reply-To: <000e01c2aad8$805f9200$0301a8c0@tom3>; from nwnetworks@dial.pipex.com on Mon, Dec 23, 2002 at 11:10:30PM -0000
Sender: owner-idr@merit.edu
Precedence: bulk
That is what I thought. It will be updated in the next revision. Andrew On Mon, Dec 23, 2002 at 11:10:30PM -0000, Tom Petch wrote: > Delivered-To: idr-outgoing@trapdoor.merit.edu > Delivered-To: idr@trapdoor.merit.edu > Delivered-To: idr@merit.edu > Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> > From: "Tom Petch" <nwnetworks@dial.pipex.com> > To: <andrewl@xix-w.bengi.exodus.net> > Cc: <idr@merit.edu>, <andrewl@cw.net> > Subject: Re: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) > Date: Mon, 23 Dec 2002 23:10:30 -0000 > X-Priority: 3 > X-MSMail-Priority: Normal > X-Mailer: Microsoft Outlook Express 4.72.2106.4 > X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 > Precedence: bulk > X-Spam-Status: No, hits=0.3 required=6.0 > tests=AWL,QUOTED_EMAIL_TEXT,SPAM_PHRASE_00_01,SUBJECT_IS_LIST, > USER_AGENT_OE > version=2.43 > X-OriginalArrivalTime: 23 Dec 2002 23:11:20.0640 (UTC) FILETIME=[9ACA4400:01C2AAD8] > > Oh dear a typo I do mean 13.5 should go. > > Tom Petch > nwnetworks@dial.pipex.com > > -----Original Message----- > From: andrewl@xix-w.bengi.exodus.net <andrewl@xix-w.bengi.exodus.net> > To: Tom Petch <nwnetworks@dial.pipex.com> > Cc: idr@merit.edu <idr@merit.edu>; andrewl@cw.net <andrewl@cw.net> > Date: 23 December 2002 19:40 > Subject: Re: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) > > > >Easy enough to do that. To clarify, in your last paragraph, do you > >mean 13.4 should go in addition to 13.5, or was the 13.4 a typo? > > > >Andrew > > > >On Mon, Dec 23, 2002 at 02:36:26PM +0000, Tom Petch wrote: > >> Date: Mon, 23 Dec 2002 14:36:26 +0000 > >> From: Tom Petch <nwnetworks@dial.pipex.com> > >> Subject: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) > >> To: andrewl@xix-w.bengi.exodus.net, idr@merit.edu > >> Cc: andrewl@cw.net > >> Reply-to: Tom Petch <nwnetworks@dial.pipex.com> > >> X-MIMEOLE: Produced By Microsoft MimeOLE V4.72.2106.4 > >> X-Mailer: Microsoft Outlook Express 4.72.2106.4 > >> X-Priority: 3 > >> X-MSMail-priority: Normal > >> > >> I think this issue SHOULD be administratively terminated. > >> > >> It relates to Connect state Event 17 (TCP connection fails) and I > am > >> credited with raising it; in fact, the issue I raised was missing > next > >> state for Active state event 17 and this has now been subsumed into > >> 13.4 (but note that 13.4 does not explicitly say Active state - I > know > >> it should because I raised that issue too). I will ensure it does > not > >> get lost from any resolution of 13.4. > >> > >> And Connect state event 17 does appear as part of issue 45 which > Siva > >> raised so I think that either way, 13.4 can go. > >> > >> Tom Petch > >> nwnetworks@dial.pipex.com > >> Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id SAA01602 for <idr-archive@nic.merit.edu>; Mon, 23 Dec 2002 18:23:11 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 4799591214; Mon, 23 Dec 2002 18:22:35 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 115F191250; Mon, 23 Dec 2002 18:22:34 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 0733491214 for <idr@trapdoor.merit.edu>; Mon, 23 Dec 2002 18:22:33 -0500 (EST) Received: by segue.merit.edu (Postfix) id D638E5DF2F; Mon, 23 Dec 2002 18:22:33 -0500 (EST) Delivered-To: idr@merit.edu Received: from demiurge.exodus.net (demiurge.exodus.net [216.32.171.82]) by segue.merit.edu (Postfix) with ESMTP id 7AC015DE64 for <idr@merit.edu>; Mon, 23 Dec 2002 18:22:33 -0500 (EST) Received: (from andrewl@localhost) by demiurge.exodus.net (8.9.3+Sun/8.9.3) id PAA24100; Mon, 23 Dec 2002 15:19:24 -0800 (PST) Date: Mon, 23 Dec 2002 15:19:24 -0800 From: andrewl@xix-w.bengi.exodus.net To: Tom Petch <nwnetworks@dial.pipex.com> Cc: idr@merit.edu, andrewl@cw.net Subject: Re: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) Message-ID: <20021223151924.G21226@demiurge.exodus.net> References: <000e01c2aad8$805f9200$0301a8c0@tom3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <000e01c2aad8$805f9200$0301a8c0@tom3>; from nwnetworks@dial.pipex.com on Mon, Dec 23, 2002 at 11:10:30PM -0000 Sender: owner-idr@merit.edu Precedence: bulk That is what I thought. It will be updated in the next revision. Andrew On Mon, Dec 23, 2002 at 11:10:30PM -0000, Tom Petch wrote: > Delivered-To: idr-outgoing@trapdoor.merit.edu > Delivered-To: idr@trapdoor.merit.edu > Delivered-To: idr@merit.edu > Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> > From: "Tom Petch" <nwnetworks@dial.pipex.com> > To: <andrewl@xix-w.bengi.exodus.net> > Cc: <idr@merit.edu>, <andrewl@cw.net> > Subject: Re: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) > Date: Mon, 23 Dec 2002 23:10:30 -0000 > X-Priority: 3 > X-MSMail-Priority: Normal > X-Mailer: Microsoft Outlook Express 4.72.2106.4 > X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 > Precedence: bulk > X-Spam-Status: No, hits=0.3 required=6.0 > tests=AWL,QUOTED_EMAIL_TEXT,SPAM_PHRASE_00_01,SUBJECT_IS_LIST, > USER_AGENT_OE > version=2.43 > X-OriginalArrivalTime: 23 Dec 2002 23:11:20.0640 (UTC) FILETIME=[9ACA4400:01C2AAD8] > > Oh dear a typo I do mean 13.5 should go. > > Tom Petch > nwnetworks@dial.pipex.com > > -----Original Message----- > From: andrewl@xix-w.bengi.exodus.net <andrewl@xix-w.bengi.exodus.net> > To: Tom Petch <nwnetworks@dial.pipex.com> > Cc: idr@merit.edu <idr@merit.edu>; andrewl@cw.net <andrewl@cw.net> > Date: 23 December 2002 19:40 > Subject: Re: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) > > > >Easy enough to do that. To clarify, in your last paragraph, do you > >mean 13.4 should go in addition to 13.5, or was the 13.4 a typo? > > > >Andrew > > > >On Mon, Dec 23, 2002 at 02:36:26PM +0000, Tom Petch wrote: > >> Date: Mon, 23 Dec 2002 14:36:26 +0000 > >> From: Tom Petch <nwnetworks@dial.pipex.com> > >> Subject: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) > >> To: andrewl@xix-w.bengi.exodus.net, idr@merit.edu > >> Cc: andrewl@cw.net > >> Reply-to: Tom Petch <nwnetworks@dial.pipex.com> > >> X-MIMEOLE: Produced By Microsoft MimeOLE V4.72.2106.4 > >> X-Mailer: Microsoft Outlook Express 4.72.2106.4 > >> X-Priority: 3 > >> X-MSMail-priority: Normal > >> > >> I think this issue SHOULD be administratively terminated. > >> > >> It relates to Connect state Event 17 (TCP connection fails) and I > am > >> credited with raising it; in fact, the issue I raised was missing > next > >> state for Active state event 17 and this has now been subsumed into > >> 13.4 (but note that 13.4 does not explicitly say Active state - I > know > >> it should because I raised that issue too). I will ensure it does > not > >> get lost from any resolution of 13.4. > >> > >> And Connect state event 17 does appear as part of issue 45 which > Siva > >> raised so I think that either way, 13.4 can go. > >> > >> Tom Petch > >> nwnetworks@dial.pipex.com > >> Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id SAA01238 for <idr-archive@nic.merit.edu>; Mon, 23 Dec 2002 18:11:29 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 6CB8791231; Mon, 23 Dec 2002 18:11:10 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 3C96291250; Mon, 23 Dec 2002 18:11:10 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id F417891231 for <idr@trapdoor.merit.edu>; Mon, 23 Dec 2002 18:11:08 -0500 (EST) Received: by segue.merit.edu (Postfix) id D60315DF18; Mon, 23 Dec 2002 18:11:08 -0500 (EST) Delivered-To: idr@merit.edu Received: from nemesis.systems.pipex.net (nemesis.systems.pipex.net [62.241.160.8]) by segue.merit.edu (Postfix) with ESMTP id A57895DF03 for <idr@merit.edu>; Mon, 23 Dec 2002 18:11:08 -0500 (EST) Received: from tom3 (userbm19.uk.uudial.com [62.188.144.225]) by nemesis.systems.pipex.net (Postfix) with SMTP id 9552716007FAE; Mon, 23 Dec 2002 23:11:01 +0000 (GMT) Message-ID: <000e01c2aad8$805f9200$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: <andrewl@xix-w.bengi.exodus.net> Cc: <idr@merit.edu>, <andrewl@cw.net> Subject: Re: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) Date: Mon, 23 Dec 2002 23:10:30 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Oh dear a typo I do mean 13.5 should go. Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: andrewl@xix-w.bengi.exodus.net <andrewl@xix-w.bengi.exodus.net> To: Tom Petch <nwnetworks@dial.pipex.com> Cc: idr@merit.edu <idr@merit.edu>; andrewl@cw.net <andrewl@cw.net> Date: 23 December 2002 19:40 Subject: Re: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) >Easy enough to do that. To clarify, in your last paragraph, do you >mean 13.4 should go in addition to 13.5, or was the 13.4 a typo? > >Andrew > >On Mon, Dec 23, 2002 at 02:36:26PM +0000, Tom Petch wrote: >> Date: Mon, 23 Dec 2002 14:36:26 +0000 >> From: Tom Petch <nwnetworks@dial.pipex.com> >> Subject: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) >> To: andrewl@xix-w.bengi.exodus.net, idr@merit.edu >> Cc: andrewl@cw.net >> Reply-to: Tom Petch <nwnetworks@dial.pipex.com> >> X-MIMEOLE: Produced By Microsoft MimeOLE V4.72.2106.4 >> X-Mailer: Microsoft Outlook Express 4.72.2106.4 >> X-Priority: 3 >> X-MSMail-priority: Normal >> >> I think this issue SHOULD be administratively terminated. >> >> It relates to Connect state Event 17 (TCP connection fails) and I am >> credited with raising it; in fact, the issue I raised was missing next >> state for Active state event 17 and this has now been subsumed into >> 13.4 (but note that 13.4 does not explicitly say Active state - I know >> it should because I raised that issue too). I will ensure it does not >> get lost from any resolution of 13.4. >> >> And Connect state event 17 does appear as part of issue 45 which Siva >> raised so I think that either way, 13.4 can go. >> >> Tom Petch >> nwnetworks@dial.pipex.com >> Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id OAA24019 for <idr-archive@nic.merit.edu>; Mon, 23 Dec 2002 14:40:55 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id F1ED79124D; Mon, 23 Dec 2002 14:40:37 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id B97589124E; Mon, 23 Dec 2002 14:40:36 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id C9E269124D for <idr@trapdoor.merit.edu>; Mon, 23 Dec 2002 14:40:35 -0500 (EST) Received: by segue.merit.edu (Postfix) id B107D5DDB8; Mon, 23 Dec 2002 14:40:35 -0500 (EST) Delivered-To: idr@merit.edu Received: from demiurge.exodus.net (demiurge.exodus.net [216.32.171.82]) by segue.merit.edu (Postfix) with ESMTP id 54B5F5DDA1 for <idr@merit.edu>; Mon, 23 Dec 2002 14:40:35 -0500 (EST) Received: (from andrewl@localhost) by demiurge.exodus.net (8.9.3+Sun/8.9.3) id LAA21412; Mon, 23 Dec 2002 11:37:30 -0800 (PST) Date: Mon, 23 Dec 2002 11:37:30 -0800 From: andrewl@xix-w.bengi.exodus.net To: Tom Petch <nwnetworks@dial.pipex.com> Cc: idr@merit.edu, andrewl@cw.net Subject: Re: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) Message-ID: <20021223113730.D21226@demiurge.exodus.net> References: <003701c2aa95$2d8a90e0$0301a8c0@tom3> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <003701c2aa95$2d8a90e0$0301a8c0@tom3>; from nwnetworks@dial.pipex.com on Mon, Dec 23, 2002 at 02:36:26PM +0000 Sender: owner-idr@merit.edu Precedence: bulk Easy enough to do that. To clarify, in your last paragraph, do you mean 13.4 should go in addition to 13.5, or was the 13.4 a typo? Andrew On Mon, Dec 23, 2002 at 02:36:26PM +0000, Tom Petch wrote: > Date: Mon, 23 Dec 2002 14:36:26 +0000 > From: Tom Petch <nwnetworks@dial.pipex.com> > Subject: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) > To: andrewl@xix-w.bengi.exodus.net, idr@merit.edu > Cc: andrewl@cw.net > Reply-to: Tom Petch <nwnetworks@dial.pipex.com> > X-MIMEOLE: Produced By Microsoft MimeOLE V4.72.2106.4 > X-Mailer: Microsoft Outlook Express 4.72.2106.4 > X-Priority: 3 > X-MSMail-priority: Normal > > I think this issue SHOULD be administratively terminated. > > It relates to Connect state Event 17 (TCP connection fails) and I am > credited with raising it; in fact, the issue I raised was missing next > state for Active state event 17 and this has now been subsumed into > 13.4 (but note that 13.4 does not explicitly say Active state - I know > it should because I raised that issue too). I will ensure it does not > get lost from any resolution of 13.4. > > And Connect state event 17 does appear as part of issue 45 which Siva > raised so I think that either way, 13.4 can go. > > Tom Petch > nwnetworks@dial.pipex.com > Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id KAA16266 for <idr-archive@nic.merit.edu>; Mon, 23 Dec 2002 10:40:03 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id C286191245; Mon, 23 Dec 2002 10:39:36 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 9235C91248; Mon, 23 Dec 2002 10:39:36 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 536F591245 for <idr@trapdoor.merit.edu>; Mon, 23 Dec 2002 10:39:35 -0500 (EST) Received: by segue.merit.edu (Postfix) id 394705DE40; Mon, 23 Dec 2002 10:39:35 -0500 (EST) Delivered-To: idr@merit.edu Received: from workhorse.fictitious.org (workhorse.fictitious.org [209.150.1.230]) by segue.merit.edu (Postfix) with ESMTP id 419BE5DE28 for <idr@merit.edu>; Mon, 23 Dec 2002 10:39:34 -0500 (EST) Received: from workhorse.fictitious.org (localhost.fictitious.org [127.0.0.1]) by workhorse.fictitious.org (8.9.3/8.9.3) with ESMTP id KAA96263; Mon, 23 Dec 2002 10:38:52 -0500 (EST) (envelope-from curtis@workhorse.fictitious.org) Message-Id: <200212231538.KAA96263@workhorse.fictitious.org> To: danny@tcb.net Cc: Mareline Sheldon <marelines@yahoo.com>, idr@merit.edu Reply-To: curtis@fictitious.org Subject: Re: Max nos of routes in FIB In-reply-to: Your message of "Sun, 22 Dec 2002 15:11:12 MST." <20021222221117.0D03355FCC@nomad.tcb.net> Date: Mon, 23 Dec 2002 10:38:52 -0500 From: Curtis Villamizar <curtis@fictitious.org> Sender: owner-idr@merit.edu Precedence: bulk In message <20021222221117.0D03355FCC@nomad.tcb.net>, Danny McPherson writes: > > > As i understand we can have some 60-100K routes in the BGP table in the > > core. I wanted an estimate as to how many of these are actually injected > > into the forwarding table? Are all these 80K+ routes actually inserted > > into the forwarding table also? > > Actually, I've seen 2+ million BGP paths (sum of all the routes in the > Adj-RIBs-In) in the core of some large networks, with ~115-160k unique > prefixes in the Loc-RIB (and FIB, subsequently). The CIDR report is > teetering around 117k unique prefixes, as of late, I believe. > > Of course, with the add-paths draft it could be be more, with the > bounded-longest-match it could be less. > > Of course, you could do clever things when populating the FIB to > reduce the size, but that's an implementation issue. > > -danny I think the question was how big does the FIB have to be and for a few ISPs with lots of internal unaggregated routes that get aggregated on the way out its around 2m+ paths (Adj-In entries) and around 200k routes (FIB entries), maybe a bit more. The number of routes in the CIDR reports represent routes after aggregation exported to a peer. Routers which are being designed so as not to become obsolete quickly have to set targets quite a bit higher. Curtis Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id KAA15119 for <idr-archive@nic.merit.edu>; Mon, 23 Dec 2002 10:10:19 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 3FE6B91246; Mon, 23 Dec 2002 10:09:19 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id E753991245; Mon, 23 Dec 2002 10:09:18 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id D035391246 for <idr@trapdoor.merit.edu>; Mon, 23 Dec 2002 10:09:15 -0500 (EST) Received: by segue.merit.edu (Postfix) id B09B95DF54; Mon, 23 Dec 2002 10:09:15 -0500 (EST) Delivered-To: idr@merit.edu Received: from colossus.systems.pipex.net (colossus.systems.pipex.net [62.241.160.73]) by segue.merit.edu (Postfix) with ESMTP id 7F8FF5DE40 for <idr@merit.edu>; Mon, 23 Dec 2002 10:09:15 -0500 (EST) Received: from tom3 (userbo34.uk.uudial.com [62.188.145.184]) by colossus.systems.pipex.net (Postfix) with SMTP id 1645316000558; Mon, 23 Dec 2002 15:09:14 +0000 (GMT) Message-ID: <003901c2aa95$2f143e20$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: <andrewl@xix-w.bengi.exodus.net>, "idr" <idr@merit.edu> Cc: <andrewl@cw.net> Subject: Issue 26 BGP Base Draft - Issue List v2.0 (Draft-19) Date: Mon, 23 Dec 2002 15:06:42 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Issue 26 discusses adding a new event or events for start with bgp-stop flap option set. But if we add an event, then we must add and agree on actions for all six existing states so I think to say that adding a new event settle things might be naive. If we do add 3) Automatic start with passive TCP establishment and bgp_stop_flap option set which I understand is Sue's resolution, then for Idle state the actions are straightforward but for the other five, is the event completely ignored? If so, does it mean that the passive flag and the bgp_stop_flap option are ignored and we carry on as if we were when we were started which may have been without them. Or is the fact of starting ignored but the flags remain set and so color the effect of other events? Needs defining. Tom Petch nwnetworks@dial.pipex.com Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id KAA15099 for <idr-archive@nic.merit.edu>; Mon, 23 Dec 2002 10:09:57 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 3C31691230; Mon, 23 Dec 2002 10:09:18 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id C970091248; Mon, 23 Dec 2002 10:09:17 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 8573791230 for <idr@trapdoor.merit.edu>; Mon, 23 Dec 2002 10:09:14 -0500 (EST) Received: by segue.merit.edu (Postfix) id 6DBBF5DF54; Mon, 23 Dec 2002 10:09:14 -0500 (EST) Delivered-To: idr@merit.edu Received: from colossus.systems.pipex.net (colossus.systems.pipex.net [62.241.160.73]) by segue.merit.edu (Postfix) with ESMTP id 308C55DE40 for <idr@merit.edu>; Mon, 23 Dec 2002 10:09:14 -0500 (EST) Received: from tom3 (userbo34.uk.uudial.com [62.188.145.184]) by colossus.systems.pipex.net (Postfix) with SMTP id D534B16000298; Mon, 23 Dec 2002 15:09:12 +0000 (GMT) Message-ID: <003801c2aa95$2e5b9c80$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: <andrewl@xix-w.bengi.exodus.net>, <idr@merit.edu> Cc: <andrewl@cw.net> Subject: Issue 45 BGP Base Draft - Issue List v2.0 (Draft-19) Date: Mon, 23 Dec 2002 14:51:56 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Issue 45 is event 17 (TCP connection fails) in Connect state which wants a textual description to reflect Jeff's statement. I propose If the TCP connection fails[Event 17] and the Open Delay timer is running, the local system: - restarts the connect retry timer, - clears the Open Delay timer - continues to listen for a connection that may be initiated by the remote BGP peer, and - changes its state to Active. If the TCP connection fails [Event17] and the Open Delay timer is not running, the local system: - drops the TCP connection, - releases all BGP resources, - sets ConnectRetryCnt (the connect retry count) to zero - resets the connect retry timer (sets to zero), and - goes to Idle state. to replace If the TCP connection fails (timeout or disconnect) [Event17], the local system: - restarts the connect retry timer, - continues to listen for a connection that may be initiated by the remote BGP peer, and - changes its state to Active. Tom Petch nwnetworks@dial.pipex.com Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id KAA15085 for <idr-archive@nic.merit.edu>; Mon, 23 Dec 2002 10:09:38 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 9933091226; Mon, 23 Dec 2002 10:09:15 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 4515291245; Mon, 23 Dec 2002 10:09:15 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id BAC5B91226 for <idr@trapdoor.merit.edu>; Mon, 23 Dec 2002 10:09:13 -0500 (EST) Received: by segue.merit.edu (Postfix) id 9E45B5DF54; Mon, 23 Dec 2002 10:09:13 -0500 (EST) Delivered-To: idr@merit.edu Received: from colossus.systems.pipex.net (colossus.systems.pipex.net [62.241.160.73]) by segue.merit.edu (Postfix) with ESMTP id 6DBF45DE40 for <idr@merit.edu>; Mon, 23 Dec 2002 10:09:13 -0500 (EST) Received: from tom3 (userbo34.uk.uudial.com [62.188.145.184]) by colossus.systems.pipex.net (Postfix) with SMTP id 111A5160001C3; Mon, 23 Dec 2002 15:09:11 +0000 (GMT) Message-ID: <003701c2aa95$2d8a90e0$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: <andrewl@xix-w.bengi.exodus.net>, <idr@merit.edu> Cc: <andrewl@cw.net> Subject: 13.5 in BGP Base Draft - Issue List v2.0 (Draft-19) Date: Mon, 23 Dec 2002 14:36:26 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk I think this issue SHOULD be administratively terminated. It relates to Connect state Event 17 (TCP connection fails) and I am credited with raising it; in fact, the issue I raised was missing next state for Active state event 17 and this has now been subsumed into 13.4 (but note that 13.4 does not explicitly say Active state - I know it should because I raised that issue too). I will ensure it does not get lost from any resolution of 13.4. And Connect state event 17 does appear as part of issue 45 which Siva raised so I think that either way, 13.4 can go. Tom Petch nwnetworks@dial.pipex.com Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id AAA19478 for <idr-archive@nic.merit.edu>; Mon, 23 Dec 2002 00:11:44 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id A1EFC9121C; Mon, 23 Dec 2002 00:11:07 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 67A659122E; Mon, 23 Dec 2002 00:11:07 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 6598A9121C for <idr@trapdoor.merit.edu>; Mon, 23 Dec 2002 00:11:06 -0500 (EST) Received: by segue.merit.edu (Postfix) id 49D3C5DEEB; Mon, 23 Dec 2002 00:11:06 -0500 (EST) Delivered-To: idr@merit.edu Received: from nomad.tcb.net (vdsl-151-118-3-177.dnvr.uswest.net [151.118.3.177]) by segue.merit.edu (Postfix) with ESMTP id 1B0A15DEE7 for <idr@merit.edu>; Mon, 23 Dec 2002 00:11:06 -0500 (EST) Received: by nomad.tcb.net (Postfix, from userid 500) id 0D03355FCC; Sun, 22 Dec 2002 15:11:17 -0700 (MST) Received: from nomad.tcb.net (localhost [127.0.0.1]) by nomad.tcb.net (Postfix) with ESMTP id 0A2073E83; Sun, 22 Dec 2002 15:11:17 -0700 (MST) X-Mailer: exmh version 2.5 07/13/2001 with nmh-1.0.4 To: Mareline Sheldon <marelines@yahoo.com> Cc: idr@merit.edu From: Danny McPherson <danny@tcb.net> Reply-To: danny@tcb.net Subject: Re: Max nos of routes in FIB Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 22 Dec 2002 15:11:12 -0700 Message-Id: <20021222221117.0D03355FCC@nomad.tcb.net> Sender: owner-idr@merit.edu Precedence: bulk > As i understand we can have some 60-100K routes in the BGP table in the > core. I wanted an estimate as to how many of these are actually injected > into the forwarding table? Are all these 80K+ routes actually inserted > into the forwarding table also? Actually, I've seen 2+ million BGP paths (sum of all the routes in the Adj-RIBs-In) in the core of some large networks, with ~115-160k unique prefixes in the Loc-RIB (and FIB, subsequently). The CIDR report is teetering around 117k unique prefixes, as of late, I believe. Of course, with the add-paths draft it could be be more, with the bounded-longest-match it could be less. Of course, you could do clever things when populating the FIB to reduce the size, but that's an implementation issue. -danny Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id XAA17431 for <idr-archive@nic.merit.edu>; Sun, 22 Dec 2002 23:35:24 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id F341F91201; Sun, 22 Dec 2002 23:35:06 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id BF0BF9121C; Sun, 22 Dec 2002 23:35:05 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id B6ECE91201 for <idr@trapdoor.merit.edu>; Sun, 22 Dec 2002 23:35:04 -0500 (EST) Received: by segue.merit.edu (Postfix) id A36835DE5B; Sun, 22 Dec 2002 23:35:04 -0500 (EST) Delivered-To: idr@merit.edu Received: from web20304.mail.yahoo.com (web20304.mail.yahoo.com [216.136.226.85]) by segue.merit.edu (Postfix) with SMTP id 05C5D5DE44 for <idr@merit.edu>; Sun, 22 Dec 2002 23:35:04 -0500 (EST) Message-ID: <20021223043503.42765.qmail@web20304.mail.yahoo.com> Received: from [165.213.1.1] by web20304.mail.yahoo.com via HTTP; Sun, 22 Dec 2002 20:35:03 PST Date: Sun, 22 Dec 2002 20:35:03 -0800 (PST) From: Mareline Sheldon <marelines@yahoo.com> Subject: Max nos of routes in FIB To: idr@merit.edu MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="0-500218568-1040618103=:42697" Sender: owner-idr@merit.edu Precedence: bulk --0-500218568-1040618103=:42697 Content-Type: text/plain; charset=us-ascii Hello, As i understand we can have some 60-100K routes in the BGP table in the core. I wanted an estimate as to how many of these are actually injected into the forwarding table? Are all these 80K+ routes actually inserted into the forwarding table also? Regards, Mareline S. --------------------------------- Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now --0-500218568-1040618103=:42697 Content-Type: text/html; charset=us-ascii <P>Hello,</P> <P>As i understand we can have some 60-100K routes in the BGP table in the core. I wanted an estimate as to how many of these are actually injected into the forwarding table? Are all these 80K+ routes actually inserted into the forwarding table also?</P> <P>Regards,</P> <P>Mareline S.</P> <P> </P> <P> </P><p><br><hr size=1>Do you Yahoo!?<br> <a href="http://rd.yahoo.com/mail/mailsig/*http://mailplus.yahoo.com">Yahoo! Mail Plus</a> - Powerful. Affordable. <a href="http://rd.yahoo.com/mail/mailsig/*http://mailplus.yahoo.com">Sign up now</a> --0-500218568-1040618103=:42697-- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id DAA11705 for <idr-archive@nic.merit.edu>; Sun, 22 Dec 2002 03:34:59 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 5397691217; Sun, 22 Dec 2002 03:34:35 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 257B591221; Sun, 22 Dec 2002 03:34:35 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 1B4ED91217 for <idr@trapdoor.merit.edu>; Sun, 22 Dec 2002 03:34:34 -0500 (EST) Received: by segue.merit.edu (Postfix) id EF6E75DE17; Sun, 22 Dec 2002 03:34:33 -0500 (EST) Delivered-To: idr@merit.edu Received: from smtp.cwnt.com (smtp.cwnt.com [192.116.246.129]) by segue.merit.edu (Postfix) with ESMTP id 3956F5DD9F for <idr@merit.edu>; Sun, 22 Dec 2002 03:34:32 -0500 (EST) X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 content-class: urn:content-classes:message Subject: Diff between -17 and -18 draft MIME-Version: 1.0 Content-Type: text/plain; charset="windows-1255" Date: Sun, 22 Dec 2002 10:34:25 +0200 Message-ID: <C7DF4400240AFB4095D98C7C6EC2A34A5DE2FC@bart.cwnt.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: Diff between -17 and -18 draft Thread-Index: AcKplO9ZrZWqCQAFRCWzwbeKbOCwgQ== From: "Amir Hermelin" <amir@cwnt.com> To: <yakov@juniper.net>, <tli@procket.com>, <skh@nexthop.com> Cc: <idr@merit.edu> Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id DAA11705 Hi, can someone point out the differences between -17 and -18 draft, either in: * short titles (sections would be helpful), or * subject of the message sent in the past to the wg, in case I missed it. any help would be greatly appreciated. I also suggest that, the documents being the length they are, you include a permanent section comparing differences to the last version of the draft (as you do with the RFCs). Thanks in advance. -- Amir Hermelin <mailto:amir@cwnt.com> Charlotte's Web Networks Inc. <http://www.cwnt.com> Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id SAA24920 for <idr-archive@nic.merit.edu>; Fri, 20 Dec 2002 18:26:38 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 0A5FE912F0; Fri, 20 Dec 2002 18:25:57 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id A963D912F1; Fri, 20 Dec 2002 18:25:56 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id C86A2912F0 for <idr@trapdoor.merit.edu>; Fri, 20 Dec 2002 18:25:51 -0500 (EST) Received: by segue.merit.edu (Postfix) id 97CB15E075; Fri, 20 Dec 2002 18:25:51 -0500 (EST) Delivered-To: idr@merit.edu Received: from demiurge.exodus.net (demiurge.exodus.net [216.32.171.82]) by segue.merit.edu (Postfix) with ESMTP id 3EF205DFFA for <idr@merit.edu>; Fri, 20 Dec 2002 18:25:50 -0500 (EST) Received: (from andrewl@localhost) by demiurge.exodus.net (8.9.3+Sun/8.9.3) id PAA02338; Fri, 20 Dec 2002 15:22:49 -0800 (PST) Date: Fri, 20 Dec 2002 15:22:49 -0800 From: andrewl@xix-w.bengi.exodus.net To: idr@merit.edu Cc: andrewl@cw.net Subject: BGP Base Draft - Issue List v2.0 (Draft-19) Message-ID: <20021220152249.D29428@demiurge.exodus.net> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="5vNYLRcllDrimb99" Content-Disposition: inline User-Agent: Mutt/1.2.5i Sender: owner-idr@merit.edu Precedence: bulk --5vNYLRcllDrimb99 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Greetings, Attached is v2.0 of the issues list. This version covers the issues that have been raised in various discussions about draft-19. As always, mistakes are mine, please let me know if I have missed anything. Andrew --5vNYLRcllDrimb99 Content-Type: text/plain; charset=us-ascii Content-Disposition: attachment; filename="Issue_List-v2.0.txt" 2002-12-09 v2.0 This is version 2.0 of this list. All the 2.x versions pertain to draft-ietf-idr-bgp4-18.txt. The revisions discussed here are targeted to be incorporated into the -19 version of this document. Since late August 2002, there has been a push to get any and all issues with the base spec. resolved so the IDR group can move this draft forward to the IESG. This is a list of the issues that have been brought up regarding the base draft, and the current working group consensus on them. All mistakes are mine, please email me at andrewl@cw.net with corrections. Please include the number and the title of the issue in the subject lines of email discussing that issue. It will help in keeping track. N.B. There is no rhyme or reason to the numbering scheme other than unique tags to address the issues. ============================================================================ Table of Contents ============================================================================ 1) Reference to RFC 1772 2) MUST/SHOULD Capitalization 3) Fix Update Error Subcode 7 -- accidently removed. 4) Section 5.1.4 - Editoral Comment 5) Section 9.1 - Change "all peers" to "peers" 6) AS Loop Detection & Implicit Withdraws 7) Standardize FSM Timer Descriptions 8) FSM MIB enumerations 9) Make "delete routes" language consistant 10) Correct OpenSent and OpenConfirm delete wording 11) Incorrect next state when the delay open timer expires. 12) Entering OpenConfirm / Adding "Stop OpenDelay" action 13) FSM Missing Next States 13.1) FSM Missing Next States - Event 15 or 16 (Connect State) 13.2) FSM Missing Next States - Event 14 (Connect State) 13.3) FSM Missing Next States - Event 15 or 16 (Active State) 13.4) FSM Missing Next States - Event 13-17 (TCP Connection) 13.5) FSM Missing Next States - Event 17 (Connect State) 13.6) FSM Missing Next States - Event 18 (Open Confirm) 14) FSM - Peer Oscillation Damping 15) FSM - Consistant FSM Event Names 16) Many Editorial Comments 17) Section 3, Page 8, Paragraph 3 - Obsolete? 18) MED Removal Text 19) Security Considerations 20) Peer Oscillation Damping 21) Session Attributes - IdleHold Timer 22) Specify New Attributes (Accept Connections/Peer Oscillation Damping) 23) Event1/Event2 Clean Up 24) Events 3, 5, 6 & 7 Give Examples 25) Event 4 & 5 Session Initiation Text 26) Event 4 & 5 - bgp_stop_flap option 27) Event 5 Clarification 28) Timer Events Definition - Make Consistant 29) Event 8 - Clean Up 30) Hold Timer - Split? 31) OpenDelay Timer Definition 32) Definition of TCP Connection Accept (Event 13) 33) Event 13 & 14 - Valid Addresses & Ports 34) Event 17 - TCP Connection Fails to TCP Connection Termination 35) Making Definition Style Consistant 36) Event 19 - Definition Cleanup 37) Event 22 - Cleanup 38) FSM Description - ConnectRetry Count 39) Handling Event 7 (Auto Stop) to Idle State processing 40) Clearing the Connection Retry Timer 41) Handling of Event 14 in the Connect State 42) Handling events 20, 21 in the Connect State and Active State 43) Handling the default events in the Connect state 44) Handling Event 23 in Connect and OpenSent 45) Event 17 in the Connect state 46) Handling of Event 17 in Active state 47) Handling of Event 19 in Active state 48) Handling of Event 2 in Active state 49) Default Event handling in Active state 50) Clearing Hold timer in OpenSent, OpenConfirm and Established State 51) Clearing Keepalive timer in OpenConfirm and Established State 52) Handling Event 18 in the OpenSent state (Keepalive Timer) 53) Established State MIB 54) State impact of not supporting Optional Events 55) New DelayOpen State ============================================================================ Issues with Consensus ============================================================================ 1) Reference to RFC 1772 2) MUST/SHOULD Capitalization 3) Fix Update Error Subcode 7 -- accidently removed. 4) Section 5.1.4 - Editoral Comment 5) Section 9.1 - Change "all peers" to "peers" 6) AS Loop Detection & Implicit Withdraws 7) Standardize FSM Timer Descriptions 8) FSM MIB enumerations 9) Make "delete routes" language consistant 10) Correct OpenSent and OpenConfirm delete wording 11) Incorrect next state when the delay open timer expires. 13.1) FSM Missing Next States - Event 15 or 16 (Connect State) 13.3) FSM Missing Next States - Event 15 or 16 (Active State) 14) FSM - Peer Oscillation Damping 16) Many Editorial Comments 19) Security Considerations 20) Peer Oscillation Damping 23) Event1/Event2 Clean Up 25) Event 4 & 5 Session Initiation Text 27) Event 5 Clarification 28) Timer Events Definition - Make Consistant 29) Event 8 - Clean Up 30) Hold Timer - Split? 31) OpenDelay Timer Definition 32) Definition of TCP Connection Accept (Event 13) 36) Event 19 - Definition Cleanup 37) Event 22 - Cleanup 38) FSM Description - ConnectRetry Count 39) Handling Event 7 (Auto Stop) to Idle State processing 46) Handling of Event 17 in Active state 49) Default Event handling in Active state 50) Clearing Hold timer in OpenSent, OpenConfirm and Established State 51) Clearing Keepalive timer in OpenConfirm and Established State 53) Established State MIB 55) New DelayOpen State ============================================================================ Issues WITHOUT Consensus ============================================================================ 12) Entering OpenConfirm / Adding "Stop OpenDelay" action 13) FSM Missing Next States 13.2) FSM Missing Next States - Event 14 (Connect State) 13.4) FSM Missing Next States - Event 13-17 (TCP Connection) 13.5) FSM Missing Next States - Event 17 (Connect State) 13.6) FSM Missing Next States - Event 18 (Open Confirm) 15) FSM - Consistant FSM Event Names 17) Section 3, Page 8, Paragraph 3 - Obsolete? 18) MED Removal Text 21) Session Attributes - IdleHold Timer 22) Specify New Attributes (Accept Connections/Peer Oscillation Damping) 24) Events 3, 5, 6 & 7 Give Examples 26) Event 4 & 5 - bgp_stop_flap option 33) Event 13 & 14 - Valid Addresses & Ports 34) Event 17 - TCP Connection Fails to TCP Connection Termination 35) Making Definition Style Consistant 40) Clearing the Connection Retry Timer 41) Handling of Event 14 in the Connect State 42) Handling events 20, 21 in the Connect State and Active State 43) Handling the default events in the Connect state 44) Handling Event 23 in Connect and OpenSent 45) Event 17 in the Connect state 47) Handling of Event 19 in Active state 48) Handling of Event 2 in Active state 52) Handling Event 18 in the OpenSent state (Keepalive Timer) 54) State impact of not supporting Optional Events ---------------------------------------------------------------------------- 1) Reference to RFC 1772 ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: Proposed changing RFC 1772 reference, since that document should be updated. Discussion: Jeff proposed that we reconsider referencing RFC 1772, since that document should be updated. Yakov pointed out that this is a non-normative reference and can just be left as is. Jeff agreed that this wasn't a big deal. We are at consensus to leave things as they are. This was discussed in the "-18 last call comments" thread. ---------------------------------------------------------------------------- 2) MUST/SHOULD Capitalization ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Capitalize MUST/SHOULD where appropate. Discussion: Jeff brought this up, and Yakov reponded asking that he point out specific instances where this is needed. Jeff said he would do so, given some time. Yakov later replied that this would be fixed in the -19 version. This was discussed in the "18 last call comments" thread. This was also brought up in the "proxy: comments on draft -18" thread. ---------------------------------------------------------------------------- 3) Fix Update Error Subcode 7 -- accidently removed. ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Add error subcode 7 back in, it looks like it was inadvertently removed. Add deprication text to Open Message Error subcode 5. Discussion: Jeff supplied: Update message error subcode 7 is removed. Especially in -18, it looks like an editing mistake based on where it would fall in the editing.. Yakov mentioned that this is addressed in Appendix A. Jeff replied: What I would like to see is something like this: 6 - Invalid ORIGIN Attribute 7 - [Deprecated - See Appendix A] 8 - Invalid NEXT_HOP Attribute As it stands, 7 lies on a page boundary and looks like it got clipped by the roff. Yakov agreed, and also said he would add similar text for Open Message Error subcode 5. This was discussed in the "18 last call comments" thread. ---------------------------------------------------------------------------- 4) Section 5.1.4 - Editoral Comment ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Fix "restricts" to "RESTRICTIONS" Discussion: Jeff proposed an editorial fix. This is agreed to. This was discussed in the "-18 last call comments" thread. ---------------------------------------------------------------------------- 5) Section 9.1 - Change "all peers" to "peers" ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Section 9.1 - Change "all peers" to "peers" Discussion: Jeff proposed: 9.1: The output of the Decision Process is the set of routes that will be advertised to (delete all) peers; the selected routes will be stored in the local speaker's Adj-RIB-Out according to policy. The previous wording implied that routes in the LocRib MUST be placed in the adj-rib-out. Yakov agreed, this fix will be in the next revision. This was discussed in the "-18 last call comments" thread. ---------------------------------------------------------------------------- 6) AS Loop Detection & Implicit Withdraws ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Update the text to reflect the AS Loop detection should be done in the BGP decision process. Discussion: John brought this up, and suggested: I have one further comment just in case it's not perfectly obvious to everyone, which is that "ignore the UPDATE" is not strictly the action you take when receiving a looped update. Rather, you treat it as an implicit withdraw, i.e. you process it as any other update but treat the contained NLRI as unfeasible. I was going to write that this is sufficiently clear from the spec, but I regret to say that it isn't. Here is the fourth paragraph of section 9: The information carried by the AS_PATH attribute is checked for AS loops. AS loop detection is done by scanning the full AS path (as specified in the AS_PATH attribute), and checking that the autonomous system number of the local system does not appear in the AS path. If the autonomous system number appears in the AS path the route may be stored in the Adj-RIB-In, but unless the router is configured to accept routes with its own autonomous system in the AS path, the route shall not be passed to the BGP Decision Process. Operations of a router that is configured to accept routes with its own autonomous system number in the AS path are outside the scope of this document. I don't think this is quite right -- the decision process needs to be run if the looped routes had previously been advertised feasibly on the same session. This could be fixed by hacking the quoted paragraph, but it seems more straightforward to do it by removing the quoted paragraph and making the fix in 9.1.2 Phase 2 instead. This could be done by inserting the following between the third and fourth paragraphs of 9.1.2 Phase 2: If the AS_PATH attribute of a BGP route contains an AS loop, the BGP route should be excluded from the Phase 2 decision function. AS loop detection is done by scanning the full AS path (as specified in the AS_PATH attribute), and checking that the autonomous system number of the local system does not appear in the AS path. Operations of a router that is configured to accept routes with its own autonomous system number in the AS path are outside the scope of this document. Section 9.3, first bullet, also addresses this topic, but I don't think it's sufficient. Yakov agreed that this was a change for the better and will include this in the next revision. We are at consensus on this issue. This is discussed in the "-18 last call comments" thread. ---------------------------------------------------------------------------- 7) Standardize FSM Timer Descriptions ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Standardize the state descriptions on those listed in the discussion section of this issue. Discussion: Tom proposed: I think a standard description would serve us better instead of using the following different ways (which I take all to refer to the same entity): delayBGP open timer BGP delay open timer BGP open delay timer delay open timer BGP delay timer I suggest Open Delay timer (with those capitals) I believe that the corresponding flag is consistently referred to (apart from the capitalisation) as Delay Open flag Yakov agreed with this suggestion, no one else disagreed, we are at consensus. This was discussed in the "BGP18-FSM-terminology" thread. ---------------------------------------------------------------------------- 8) FSM MIB enumerations ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Move MIB references from the base spec into the MIB document. Discussion: Tom pointed out that: The FSM makes several references to putting values into MIB objects and while some of the values are defined, eg FSM error or Hold Timer expired, I can find no definition of the following in any of the BGP documents, MIB or otherwise. connect retry expired TCP disconnect administrative down collision detect closure Call Collision cease collision detected and dump connection Administrative stop I believe an implementation needs to be told these values somewhere and that there should be a reference to that place in bgp18. Jeff replied that to make things easier, the MIB references will be removed from the base spec, and into the MIB document. This was discussed in the "WG Last Call FSM MIB enumeration" thread, and the "bgp18 WG Last Call fsm MIB objects" thread. ---------------------------------------------------------------------------- 9) Make "delete routes" language consistant ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Replace a variety of wording with "deletes all routes associated with this connection,". Discussion: Tom pointed out that we use a variety of language to say how we are going to delete routes in the FSM. He proposed that we instead use: - deletes all routes associated with this connection, This met with agreement, and will be reflected in the next version. This was discussed in the "bgp18 WG Last Call fsm delete action" thread. ---------------------------------------------------------------------------- 10) Correct OpenSent and OpenConfirm delete wording ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Remove delete wording from OpenSent and OpenConfirm states. Discussion: Venu asked why there was delete wording in the OpenSent and OpenConfirm states when a BGP speaker cannot recieve routes in these states. Jeff acknowledged that this was an error. Yakov agreed to fix the next version. This was discussed in the "bgp18 WG Last Call fsm delete action" thread. ---------------------------------------------------------------------------- 11) Incorrect next state when the delay open timer expires. ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Fix the next state. Discussion: Tom pointed out that: I believe that there is an incorrect next state when the delay open timer expires [event 12] in the Active state. The next state should be OpenSent and not OpenConfirm. OpenConfirm is for KeepAlive processing when Open messages have been sent and received. OpenSent is for Open sent and not yet received. The corresponding section in Connect state I believe is correct. Yakov agreed, and will fix this in the next revision. This was discussed in the "bgp18 WG Last CAll fsm incorrect next state" ---------------------------------------------------------------------------- 12) Entering OpenConfirm / Adding "Stop OpenDelay" action ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Add new text agreeded upon in the discussion section, resolve possible addition of Stop OpenDelay action (not state, which we agreed doesn't reflect current practice.) Discussion: This discussion began with Tom outlining these two points: When the OpenConfirm state is entered from OpenSent with the receipt of a valid open [Event 18], then a KeepAlive message is sent and the timer is started. When the OpenConfirm state is entered from Active or Connect on receipt of a valid open [Event 19], no message is sent, no timer is started. I believe this inconsistency is an error and should be corrected by adding these two actions in those two places. Sue replied: Just to clarify this comment: Event 19 = valid open with delay timer running Active = 1) awaiting TCP connection, or 2) TCP connection completed and awaiting the TCP connection with delay timer running Case 1: - should not see Event 19 In transition from Active to Open Confirm, the connection must have a TCP connection completed. Case 1 does not have this occurring, so the transition must be avoided. Case 2: - should see Event 19 - Open, Keepalive should be sent. Previous text: (Action H from FSM document) If an Open is received with the BGP Delay Open timer running, [Event 19], the local system: - clears the connect retry timer [cleared to zero), - completes the BGP initialization, - stop and clears the BGP Open Delay timer, - Sends an Open Message, - sets the hold timer to a large value (4 minutes), and - changes its state to an Open Confirm. New text: [a New Action - N-2 : N + BGP keepalive sent] If an Open is received with the BGP Delay Open Timer running [Event 19], the local system: - clear the connect retry timer [cleared to zero], - completes the BGP initialization, - stops and clears the BGP Open Delay timer, - Send an Open message, - Sends a Keepalive message, - If hold timer value is non-zero, - set keepalive timer - hold timer reset to negotiated value else if hold timer is zero, - reset the keepalive timer, and - reset the hold timer. - If the value of the autonomous system field is the same as the local Autonomous system number, set the connection status to a internal connection; otherwise it is "external". Tom and Sue discussed the OpenDelay state, and recalled that this was excluded a number of months ago as not reflecting current practice. By way of clarification, Sue added: 1) Agree, this can occur in the Active state as well as the Connect state. Will you accept the earlier text below to be inserted both places? Background: The state machine for Event 19 is: Idle Connect Active Open Open Estb Sent Confirm =============================================== Event 19 | | | | | | | next state |Idle | Open | Open | Open |Idle | Idle | | | confirm| confirm| Confirm| | | =============================================== action | V | N-2 | N-2 | N | E-1 | E | =============================================== Per the State Machine. Action v - FSM Error Action E - FSM Error, drop connection - etc, drop routes Action E-1 - FSM Error, drop connection (lots of Action N-2 (text below) Action N (text below, without sending Open) 2) Do you think that Event 19 is possible in the Open Sent state? Please answer this separately. Tom replied that: 1) yes I think the same text in both Active and Connect states is a good resolution 2) complicated. As the fsm text stands, Event 19, along with a host of others, takes us back from Open Sent to Idle (I assume on the grounds this is an error condition) which seems very reasonable. But ...in quite a few places, such as Connect state events 2, 7,8,9,10,11, 17, 18, 20 thru 27, we do not stop the OpenDelay timer when going to Idle or Active so we could then go from eg Idle with Manual start [event 1] to Connect to Open Sent all before the OpenDelay timer expires in which case event 19 can occur validly in Open Sent - obscure but possible. (This is also true with Active state and events 2, 17 and the default list at the end). But I think this is an error, and that when exiting Connect state or Active state as listed above, we should have an additional action to stop the OpenDelay timer in which case event 19 in Open Sent becomes an error condition (again). But but but as ever, I cannot speak with authority for implementations and so if implementations do not stop the OpenDelay timer when exiting as above, then Event 19 is valid in Open Sent state - obscure but possible (again). My wish is to add the extra action, stop OpenDelay timer, for the events listed above in Active and Connect states in the expectation that that is what people have or should have implemented. Tom added a reponse to Sue after some other threads have been discussed: You asked if event 19 (Open with OpenDelay timer running) was possible in OpenSent state; I gave a lengthy reply (below) to the effect yes it could because the OpenDelay timer did not always get stopped but the timer should be stopped in which case the event would not happen. Reading your responses to Siva , I see you include stopping the Open Delay timer in the action 'release all BGP resources' when going to Idle (which I missed seeing earlier in the year). That eliminates most but not all of the possibilities I mentioned. I now believe we would need to add the action 'stop OpenDelay timer' for Connect state event 17 (currently defined as going to Active) event 9 (stays in Connect state) in order to stop event 19 in Open Sent This was discussed in the "bgp 18 WG Last Call fsm missing keepalive" thread. And also in the "Event 19 in Open Sent state was Re: bgp18 WG Last Call fsm missing keepalive" thread. ---------------------------------------------------------------------------- 13) FSM Missing Next States ---------------------------------------------------------------------------- Status: See sub-issues Change: Potentially Summary: Seven sub-issues spawned to resolve each of the next-state questions. Discussion: This began with Tom pointing out 7 places where the next state was not clear. Interlaced with his comments below is the proposed text to fix the problems and the status of the issue. This conversation was started in the "bgp18 WG Last Call fsm missing next state" thread. ---------------------------------------------------------------------------- 13.1) FSM Missing Next States - Event 15 or 16 (Connect State) ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Add next state of Connect. Discussion: Tom pointed out that: Connect State: If the TCP connection succeeds [Event 15 or Event 16], the local system checks the "Delay Open Flag". If the delay Open flag is set, the local system: **enters what state Sue proposed these changes: 1) Connect State - Event 15 or Event 16 [consensus, editorial] note: The delay retry timer is utilized instead of the connect retry timer for the next two changes. Previous text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the "Delay Open Flag". If the delay open flag is set, the local system: - clears the connect retry timer, - sets the BGP open delay timer to initial value If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clear BGP Open Delay timer (set to zero), - completes the BGP initialization, - send an Open message to its peer, - sets hold timer to a large value, and - change the state to Open Sent. New text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the Delay Open flag prior to processing: If the Delay Open flag is set, the local system: - clears the connect retry timer, - sets the BGP open delay timer to initial value, and - stays in the Connect state. If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clears the BGP Delay timer (sets to zero), - completes the BGP initialization, - sends an Open message to its peer, - sets the hold timer to a large value, and - changes the state to Open Sent. Tom agreed that this was good, with the change to "Open Delay timer" as discussed in issue 7. This conversation was started in the "bgp18 WG Last Call fsm missing next state" thread. ---------------------------------------------------------------------------- 13.2) FSM Missing Next States - Event 14 (Connect State) ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Choose one of two alternatives for the next state after event 14. Possibly add a clause about peer oscillation damping. Discussion: Tom pointed out that: Connect State: If the TCP connection receives an indication that is invalid or unconfigured. [Event 14]: **enters what state Sue proposed these alternatives: 2)Connect State - Event 14 [no consensus] Current Text: If the TCP connection receives an indication that that is invalid or unconfigured [Event 14], - the TCP connection is rejected. At the very least this section needs more "word smithing", so I'd like to change it for more clarity at least. I'm not sure this represents the implementations. What I'd like to do is query the implementations to see what they do if they receive a valid TCP connection with an invalid or unconfigured peer. Two options: Alternative 1: Count it as a valid response New Text: If a TCP connection is received that has an invalid format, or an unconfigured host [Event 14], the local system: - rejects the TCP connection, - increments the connect retry counter, - performs bgp peer oscillation checks. If bgp peer oscillation checks allow for a new connection, the bgp peer - restarts the Connect retry timer with configured value, and - enters the Active state. FSM table: Idle Connect Active Open-Sent Open-Confirm Establish Event-14 ======================================================= Next state Idle | Active|Active|Open-Sent|Open-Confirm|Establish| ======================================================= action V | Y2 | L | Ignore | Ignore | Ignore | ======================================================= Alternative 2: Reject the connection and see if valid or configured one appears within the connect timer window. New Text: If a TCP connection is received that has an invalid format, or an unconfigured host [Event 14], the local system: - rejects the TCP connection, - and stays in the Connect state. FSM table: Idle Connect Active Open-Sent Open-Confirm Establish Event-14 ======================================================= Nxt state Idle |Connect|Active|Open-Sent|Open-Confirm|Establish| ======================================================= action V | L | L | Ignore | Ignore | Ignore | ======================================================= Sue then sent out a call to implementors to let the list know what they did with their FSMs. Tom replied that he agreed that we need to wait to see what the existing implementations do. He also suggested: **tp need a then clause here 'if bgp peer oscillation damping does not allow for a new connection, then the local system ???' be added before the FSM table in option 1 of the proposed text. This conversation was started in the "bgp18 WG Last Call fsm missing next state" thread. ---------------------------------------------------------------------------- 13.3) FSM Missing Next States - Event 15 or 16 (Active State) ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Add text listed in discussion. Discussion: Tom pointed out: Active State: A TCP connection succeeds [Event 15 or Event 16], the local system: process the TCP connection flags - If the BGP delay open flag is set: ** enters what state (I think this is an FSM error in TCP because it has not initiated a connection!) Sue proposed these changes: Previous text: A TCP connection succeeds [Event 15 or Event 16], the local system: process the TCP connection flags - If the BGP delay open flag is set: - clears the connect retry timer [through the following text: - and changes its state to Open Sent. New text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the "Delay Open Flag" prior to processing: If the delay open flag is set, the local system: - clears the connect retry timer, - sets the BGP open delay timer to initial value, and - stays in the Active state. If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clears the BGP Delay timer (sets to zero), - completes the BGP initialization, - sends an Open message to its peer, - sets the hold timer to a large value, and - changes the state to Open Sent. Tom agreeded with this. This conversation was started in the "bgp18 WG Last Call fsm missing next state" thread. ---------------------------------------------------------------------------- 13.4) FSM Missing Next States - Event 13-17 (TCP Connection) ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Get feedback from implementors on TCP connection events. Discussion: Tom started this by saying that: If the local system receives a valid TCP Indication [Event 13], the local system processes the TCP connection flags. ** enters what state If the local system receives a TCP indication that is invalid for this connection [Event 14]: ** enters what state Sue proposed we move this to the "fsm missing next state - Events 13-17 and the TCP connection" thread. The response in this thread was: 4) Active State, Event 13 [no consensus] 5) Active State, Event 14 [no consensus] The problem with this state is it is difficult to exactly specify without discussing the TCP Messages that FSM document covers. I'll query if the implementors require all of events 13-17 as mandatory. Sue polled the implemtentors on the list with this query: These events are described in section 8.1.3. In our discussion in January through May of 2002, many implementers mapped their implementation onto the following TCP events list in 8.1.3. Events 13 - 17 Event 13 - TCP connection indication & valid remote peer Event 14 - TCP connection indication with invalid source or destination Event 15 - TCP connection request sent (by this peer) received an Acknowledgement [ local system sent a TCP SYN, Received a TCP SYN, ACK pair back, and Sent a TCP ACK] Event 16 - TCP connection confirmed [local system received a TCP SYN, sent a TCP SYN, ACK back, and received a TCP ACK] Event 17 - TCP connections Should we have all of these states? Which implementations support all of these Events? The full FSM text was snipped here for brevity. This was started in the "bgp18 WG Last Call fsm missing next state" thread. And continued in the "fsm missing next state - Events 13-17 and the TCP connection" thread. ---------------------------------------------------------------------------- 13.5) FSM Missing Next States - Event 17 (Connect State) ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Choose one of the two options outlined in the discussion. Currently Option 2 has received the most support. Discussion: If the local system receives a TCP connection failed [Event 17] (timeout or receives connection disconnect), the local system will: ** enters what state Sue replied with this: comment: In the Active state, we may already have a connection and be awaiting the Open Delay timer. The TCP disconnect or timeout could occur in this state due to the "Open Delay Timer". If the TCP Disconnect is ignored, we could have some peer oscillation. If the we wait, then the connection retry timer needs to be kept running. The text below allows this timer. The real question is what is the status of the current implementations. I agree, the Active state and the connect state should match. Old Text: If the TCP connection fails (timeout or disconnection) [Event 17], the local system: - set TCP disconnect in the MIB reason code, - restart Connect retry timer (with initial value), - release all BGP resources, - Acknowledge the Drop of the TCP connection if TCP disconnect (FIN ACK), - Increment ConnectRetryCnt (connect retry count) by 1, and - performs the BGP peer oscillation damping process. Applicable FSM State table: FSM table old: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Active |Idle |Active | Idle |Idle | | | | | | | ========================================================= action V | Y2 | G | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= Alternative 1: FSM table new: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Active |Active |Active | Idle |Idle | | | | | | | ========================================================= action V | G | G | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= G: The local system: - restarts the connect retry timer (at intial value), - continues to listen for a connection that may be initiated by the remote peer, and - sets its next state to Active. New Text: (for Connect and Active state) If the TCP connection fails (timeout or disconnect) [Event 17], the local system: - restarts the connect retry timer, - continues to listen for a connection that may be initiated by the remote BGP peer, and - changes it state to Active. Alternative 2: FSM table new: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Idle |Idle |Active | Idle |Idle | | | | | | | ========================================================= action V | Y2 | Y2 | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= Next Text: If the location system receives a TCP connection failed [Event 17], the local system will: - increment the ConnectRetrycnt (connect retry count) by 1, - release all BGP resources associated with this connection, - perform BGP peer oscillation (if configured), and - go to Idle Y2 - is: The local system: 1) increments the ConnectRetryCnt (connect retry count) by 1, 2) releases all BGP resources associated with this connection, and 3) performs the BGP peer oscillation damping process if the damping process allows for a new connection, the local system - restarts the connect retry timer (with initial value, and - goes to Idle If the damping process does not allow for a new connection, the local system - set the flags to damp the creation of a new bgp connection until a manual start occurs, and - goes to Idle. Tom agreed with the options, and stated that he prefered option 2. Sue is also happy with option 2, if no one else chimes in. This conversation was started in the "bgp18 WG Last Call fsm missing next state" thread. ---------------------------------------------------------------------------- 13.6) FSM Missing Next States - Event 18 (Open Confirm) ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Get reponses from implementors, get collision detection text cleared up. Proposed text is below. Discussion: Tom opened this with: Open Confirm State: If the Open messages is valid [Event 18], the collision detect function is processed per section 6.8. If this connection is to be dropped due to call collision, the local system: ** enters what state Sue replied with: Here's my proposed text. Please let me know what you think. I think this is an editorial change. Old text: If the open message is valid, the collision detect function is processed per section 6.8. If this connection is to be dropped due to call collision, the local system: - sends a Notification with a Cease - resets the Connect timer (to zero), - releases all BGP resources, - Drop the TCP connection (sends a TCP FIN), - increments the ConnectRetryCnt by 1 (connect retry count), and - performs an BGP peer oscillation damping process. New text: If the open message is valid, the BGP peer connection is check to detect a collision per section 6.8. If this connection is to be dropped due to call collision, the local system: - sends a Notification with a Cease - resets the Connect timer (to zero), - releases all BGP resources, - Drop the TCP connection (sends a TCP FIN), - increments the ConnectRetryCnt by 1 (connect retry count), and - performs an BGP peer oscillation damping processing, and - enters the Idle State. notes: Collision detect impacts Open Sent, Open Confirm, and Established states. Tom replied: I am still struggling with; we are in OpenConfirm so we already have received an Open from the remote peer and Event 18 is a second Open from the same peer. Perhaps my struggle is that I think in terms of two (or more) FSM for a given IP address pair so the Open Collision detection will occur when the/an- other FSM receives a valid Open in states Active/Connect/Open Sent and will generate Event 22 into this FSM so Event 18 cannot occur. But yes, if Event 18 can occur in this FSM and this connection is to be dumped, then Idle state it should be as you suggest. I have slotted in [optionally] in front of the peer oscillation damping in your text because I think it should be optional:-) Sue replied: this mechanism allows a single fsm to handle both. 2 fsm and 1 fsm BGP FSM seem to exist. (I queried implementors a few times on this one. So, I just put in this change to provide the flexibility. Collision detect tends to give scrambled brains for most people.. As Dennis Ferguson said 2 years ago, that's the hardest part of the FSM. Sue then stated that she would query implementors to see what is being done. This conversation was started in the "bgp18 WG Last Call fsm missing next state" thread. ---------------------------------------------------------------------------- 14) FSM - Peer Oscillation Damping ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Change references to peer oscillation damping to consistant phrase: "[optionally] performs peer oscillation damping". Also remove old reference to "BGP Peer Restart Backoff Mechanisms". Discussion: Tom suggested we use consistant terminology to refer to peer osillation damping. He also pointed out a stale reference. Yakov agreed to fix both of these. ---------------------------------------------------------------------------- 15) FSM - Consistant FSM Event Names ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Proposal to make the FSM Event Names consistant. There have been no comments on this propsal. Discussion: Tom proposed that: The event name used in the FSM show much variation to the point sometimes where I am not clear that it is always the same event (eg where the event name is qualified by a subset of the possible causes). Assuming that it is, I propose the following changes to make the wording consistent, clear and concise for event names. ** denotes changed text using the convention /'old text'/'new text'/ 8. BGP Finite State machine Event1: Manual start Event2: Manual stop Event3: Automatic start **Event4: Manual start with passive TCP /estabishment/flag/ **Event5: Automatic start with passive TCP /establishment/flag/ Event6: Automatic start with bgp_stop_flap option set **Event7: Auto//matic/ stop Event8: Idle hold timer expires Event9: Connect retry timer expires **Event10: Hold time//r/ expires Event11: Keepalive timer expires Event12: Open Delay timer expires **Event13: TCP connection valid indication **Event14: TCP connection invalid indication **Event15: TCP connection request /sent received an ACK/acknowledged/ Event16: TCP connection confirmed Event17: TCP connection fails Event18: BGPOpen Event19: BGPOpen with *Open Delay timer running Event20: BGPHeaderErr Event21: BGPOpenMsgErr Event22: Open collision dump Event23: NotifMsgVerErr Event24: NotifMsg Event25: KeepAliveMsg Event26: UpdateMsg Event27: UpdateMsgErr 8.2.2 Finite State Machine Connect State: If the BGP port receives a ** valid TCP connection indication [Event 13], If the TCP connection receives **an invalid indication [Event 14]: If the TCP connection fails **/(timeout or disconnect)// [Event17] Active State: If the local system receives a **valid TCP //indication/ [Event 13], If the local system receives a TCP connection failed [Event 17] **/(timeout or receives connection disconnect)//, Open Sent: If a connection in Open Sent is determined to be the connection that must be closed, an **/administrative collision detect/Open collision dump/ [Event 22] is signaled to the state machine. If such an **/administrative collision detect dump [Event 22]/event/ is If a TCP **//connection valid/ indication [Event 13] or TCP **//connection/ request **//acknowledged/ [Event 15] Open Confirm State: ... or receives a TCP **/Disconnect// connection fails/ [Event 17] from the In the event of **/TCP establishment//TCP connection valid indication /[Event 13] .... the local system will **/issue a call/generate an Open/ collision dump [Event 22]. When the local system receives a **/call/open/ collision dump event [Event 22]/such an event/, the Established State: **/disconnect from the underlying TCP/TCP connection fails/ [Event17], it: ... it will process **/a Call/an Open/ Collision dump event[Event 22]. Notes: Event 4 title brought in line with text Event 5 title brought in line with text Event 7 title brought in line with text Event 13 title shortened to be closer to text, text brought in line Event 14 title shortened to be closer to text, text brought in line Event 15 title brought in line with text Event 17 text brought in line with title (text often introduces qualifying conditions that are too restrictive) Event 22 text brought in line with title This is discussed in the thread: "bgp18 WG Last Call fsm event names." ---------------------------------------------------------------------------- 16) Many Editorial Comments ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Many editorial suggestions, and what we are doing with them are listed below. Some issues have been broken out seperately where there is a longer discussion on them. Discussion: Alex began this by presenting comments from an anonymous reviewer, unless otherwise noted, responses are from Yakov: > Almost all of these are simple clarifications. > > Section 1, page 5: IGP definition - it's not clear from this > definition whether IBGP would be considered an IGP? any suggestion on how to improve the definition to clarify this issue would be appreciated. There was some further discussion on this and it was decided that people reading this document ought to know what an IGP is. > Section 3, page 7, para 4: Does RFC 1772 still represent the *planned* > use of BGP? Or the actual use? Or something different from actual > use? Perhaps we should just take out references to 1772. Further discussion seemed to indicate that this reference should stay. > Section 3, page 8, para 3 - "The hosts executing..." This paragraph > seems obsolete. I'll take it out. With regard to this, Siva asked if some route optimizaiton vendors rely on this. Since this wasn't resolved, it is discussed further in issue 17. > Section 4.1, page 11 - Length is in network byte order. all the encodings are in network byte order. This applies not just to the BGP spec, but to other protocols as well. This comment was made about a number of fields. It was later agreed that a reference would be made to this at the beginning of the document. > Section 4.2, page 12 - Hold Time - what does a value of zero indicate? if you read section 4.4 then you'll find that: If the negotiated Hold Time interval is zero, then periodic KEEPALIVE messages MUST NOT be sent. > Section 4.2, page 13 - BGP Identifier - network byte order? > "IP address" -> "IPv4 address" I'll put at the beginning a sentence saying that in the context of this document the term "IP address" means an IP Version 4 address. > Section 4.3, Page 14, para1, sentence 2 - "path attribute" -> "path > attributes" fixed. > Section 4.3,Page 17, NEXT_HOP: "IP address" -> "IPv4 address" > Specify that this is 4 octets. > Reference here to multi-protocol extensions for IPv6 nexthop? no. > RFC 2283 is unclear whether NEXT_HOP should always be included > when using multiprotocol extensions. Clarify this here? It is already clarified in 2283bis. > Section 4.3, Page 17/18 - MED and LocalPref: > "non-negative" -> "unsigned" for consistency with elsewhere. > (non-negative might imply values > 2^31 cannot be used). fixed. > Section 4.3, Page 19 - Prefix: "IP address" -> "IPv4 address" > Prefix: "enough trailing bits to" -> "the minimum number of > trailing bits needed to" fixed. > Section 4.4, Page 20: - "BGP does not use any TCP-based keep-alive > mechanism to determine if peers are reachable". Is it worth noting > that TCP may still timeout the connection even if TCP keepalives are > turned off? the text is fine as it is. > Section 4.4, Page 20: > KEEPALIVE message consists" -> "A KEEPALIVE message consists" fixed. > Section 5, Page 23: "The same attribute can not appear more than once > with the Path Attributes field...". Does this mean the same attribute > type, or the same attribute type and value? the former (the same attribute type). > Section 5.1 "The usage of each BGP path attributes .." -> attribute fixed. > Section 5.1.3 "IP address" -> "IPv4 address" > > "A BGP speaker must never advertise an address of a peer to that peer > as a NEXT_HOP, for a route that the speaker is originating." > suggest replace this text with: > "A route originated by a BGP speaker must never be advertised to a > peer using an address of that peer as NEXT_HOP" fixed. > Section 5.1.4: "A BGP speaker MUST IMPLEMENT a mechanism ... which > allows the MULTI_EXIT_DISC to be removed from a route." Might want to > say that this is dangerous unless you received the route from an EBGP > peer? think we should keep the text as is. > Section 5.1.5: "If it [LOCAL_PREF] is contained in an UPDATE message > that is received from an external peer, then this attribute MUST be > ignored by the receiving speaker, except for the case of BGP > Confederations [RF3065]." > - "ignored" might be taken to mean that you don't process it for > decision, but that you propagate it to internal peers. I might > write "silently removed" or something similar. I think the text is ok as is. > Section 5.1.5, para 2. "set of AS" -> "set of ASs" fixed. > Section 6.3: wrt NEXT_HOP semantic correctness: should we check that a > NEXT_HOP is not a multicast or broadcast address? I'll add to the definition of NEXT_HOP that it is a unicast address. > Section 6.3, page 32, para 7: "peer than sent" -> "peer that sent" fixed. > Section 6.3: "if any attribute appears more than once" - does this > mean the same attribute type, or the same attribute type and value? the former. > Section 6.8 "Comparing BGP identifiers is done by treating them as > (4-octet-long) unsigned integers". Need to convert to host byte order > before comparing. fixed. > Section 6.8, item 2: "closes BGP connection" -> "closes the BGP connection" > "accepts BGP connection" -> "accepts the BGP connection". fixed. > Section 9.1.2.2: item (c): in the explanation of neighborAS(n), it is > unclear for IBGP connections how to determine "the neighbor AS from > which the other IBGP speaker learned the route". If this is really > the leftmost entry in the AS path (or the local AS if the path is > empty), the spec should explicitly say so. fixed. > Section 9.1.2.2, page 63, paragraph starting "If a MULTI_EXIT_DISC > attribute is removed before..." The first sentence is pretty nearly > incomprehensible. This topic has some more discussion surrounding what text we should use to clarify this issue. This is followed up in issue 18. > Section 9.1.2.2 (d) > "d) If at least one of the candidate routes was received from an > external peer in a neighboring autonomous system, remove from con- > sideration all routes which were received from internal peers." > For consistency with (c) and clarity, this might be reworded: > "d) If any of the candidate routes was learned via EBGP, remove > from consideration all routes which were learned by IBGP." fixed. > Section 9.1.2.2 (e) > "cost (n) is better than cost (m)" > Given the definition of cost, it might be clearer to say > "cost (n) is lower than cost (m)" fixed. > Section 9.1.2.2 (g) > "neighbor address" has not been defined. I'll replace "neighbor address" with "peer address". > Section 9.2.2.2, Page 70 (AGGREGATOR) - "All AGGREGATOR attributes of > all routes to be aggregated should be ignored." > > Perhaps "ignored" is ambiguous here, and it's not clear whether > should is a SHOULD. Suggest: > > "Any AGGREGATOR attributes from the routes to be aggregated MUST > NOT be included in the aggregated route." fixed. > Section 9.3 - shouldn't this subsection be moved to the discussion of > Phase 1 or Phase 2 of the decision process? Or at least move it > before Section 9.2. I think it is fine where it is now. > Appendix E, para 2: IP precedence has been deprecated. Delete this > paragraph, or replace with appropriate diffserv codepoint. deleted. > Security Considerations: > "BGP supports the ability to authenticate BGP messages by using BGP > authentication." > This sentence should be removed, and the Authentication Information > parameter has been deprecated. Please see the recent e-mail exchange on the Security Considerations See issue 19 for more on the Security Considerations section of the draft. These topics were discussed in the "proxy: more comments on the draft -18" thread. ---------------------------------------------------------------------------- 17) Section 3, Page 8, Paragraph 3 - Obsolete? ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Do existing vendors rely on this, or no? If no, remove it. Discussion: This issue was spawned from the discussions in issue 16, specifically: Anonymous reviwer: > Section 3, page 8, para 3 - "The hosts executing..." This paragraph > seems obsolete. Yakov: I'll take it out. With regard to this, Siva asked if some route optimizaiton vendors rely on this. This was discussed in the "proxy: more commetns on draft -18" thread. ---------------------------------------------------------------------------- 18) MED Removal Text ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Select new text from the options below. Discussion: This issue is spawned from issue 16. An anonymous reviewer pointed out: > Section 9.1.2.2, page 63, paragraph starting "If a MULTI_EXIT_DISC > attribute is removed before..." The first sentence is pretty nearly > incomprehensible. Yakov replied: here is my attempt to clarify this: If a MULTI_EXIT_DISC attribute is removed before re-advertising a route into IBGP, then (prior to the removal) the MULTI_EXIT_DISC attribute may only be considered in the comparison of EBGP learned routes; the attribute is then removed, and then the remaining EBGP learned routes may be compared to the remaining IBGP learned routes, without considering the MULTI_EXIT_DISC attribute for those EBGP learned routes whose MULTI_EXIT_DISC attribute will be removed before advertising these routes to IBGP. Any further suggestions on how to improve this would be appreciated. Siva replied: How about this: % If a MULTI_EXIT_DISC attribute is removed before re-advertising a % route into IBGP, then comparison based on the MULT_EXIT_DISC % attribute may (MUST?) be performed only among the EBGP learned routes. % This comparison MUST be performed before the removal of the % MULTI_EXIT_DISC attribute. The MULT_EXIT_DISC attribute must then % be removed from those EBGP routes where such removal is required and % which are still eligible. This is followed by comparison with IBGP % learned routes. I think this reflects our objectives, which is: a) If MED is to be removed, compare EBGP routes based on the MED b) Then remove the MED c) Then do comparison with IBGP routes Andrew suggested: If a router is configured to remove a MUTLI_EXIT_DISC attribute from a route learned from EBGP, before readvertising it into IBGP the router MUST compare the route with other EBGP-learned routes before removing the MULTI_EXIT_DISC. Once this comparison is complete, the MED may be removed, and any remaining routes can be compared with IBGP routes to determine the best route. This is discussed in the "proxy: more comments on draft 18" thread. ---------------------------------------------------------------------------- 19) Security Considerations ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Fix Security Considerations section to include manditory MD5 auth and advance security considerations draft along with the base draft. Discussion: Yakov started this discussion by proposing text which would require TCP MD5 authententication for BGP implementations. This is to bring the spec in line with an IETF requirement that authentication be available. After some discussion the plan is to advance draft-ietf-idr-bgp-vuln-00.txt as Informational along with the base BGP specification. This draft will serve as the security analysis section of the base spec. This is discussed in the "revised Security Considerations section" thread. ---------------------------------------------------------------------------- 20) Peer Oscillation Damping ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: Keep the Peer Oscillation Damping reference in the specification. Discussion: This began when Siva proposed: Since this feature is going to be added in a new draft, and its addition will change the operation of the state machine, can we remove all mention of it in the state machine ? As part of this removal, can we also remove the IdleHold Timer from the FSM since it is not useful in the absence of peer oscillation damping ? The draft that describes this procedure can then describe the change in the state machine required to do this. Sue replied that: The reason we should not remove the peer oscillation damping from the state machine: 1) Deployed implementations support peer oscillation damping 2) Hooks for the additions in the FSM cannot be added later. These hooks are optional and do not need to be implemented. Siva replied: I understand. I am not trying to object to peer oscillation damping, I think it is a good idea and we have included it in our implementation as well. I was suggesting that instead of a partial description in this draft, it be completely described in the draft on peer oscillation damping. However, I do see your point, and unless there are any objections from others, I think we have consensus on this issue. This was discussed in the "Response to FSM input - Comments 1-10" thread: Comment #1. ---------------------------------------------------------------------------- 21) Session Attributes - IdleHold Timer ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Text about the IdleHold timer is on the table, we need to decide if it is what we want. The addition of the DelayBgpOpenTimer and DelayOpen Flag have not yet been addressed. Discussion: This discussion began with Siva asking: Why have a Hold Timer and a Hold Time ? Can we replace this with just Hold timer ? Can we also add the following session attributes: a) DelayBgpOpenTimer b) IdleHold Timer (in case we choose not to remove this from the bas = FSM) Can we also add the following flag to the session attributes: a) DelayOpen Flag After some discussion we have this text on the table: Event8: Idle hold timer expires Definition: An Event generated when the Idle Hold Timer expires. The Idle Hold Timer is only used when the persistent peer oscillation damping function is enabled. % % Implementations not implemented persistent % peer oscillations damping functions may not % have the Idle Hold Timer. This was discussed in the "Response to FSM input - Comments 1-10" thread: Comment #2. ---------------------------------------------------------------------------- 22) Specify New Attributes (Accept Connections/Peer Oscillation Damping) ---------------------------------------------------------------------------- Status: No Consensus Change: Potentailly Summary: Is proposed text acceptable? Discussion: This begain with Siva propsosing: Can we call these out as well: * Accept Connections from unconfigured peers (Enabled/Disabled) * Peer Oscillation Dampening (Enabled/Disabled) (In case we choose not to remove it from base spec) After some discussion we have this text on the table: The following will be added to 8.0 Optional parameters that may be supported either per connection or per implementation: 1) Delay Open flag 2) Delay Open Timer 3) Perform automatic start flag 4) Passive TCP establishment flag 5) BGP stop_peer_flag flag 6) Idle Hold timer 7) Perform automatic stop flag 8) Perform Collision detect in Establish mode flag This was discussed in the "Response to FSM input - Comments 1-10" thread: Comment #3. ---------------------------------------------------------------------------- 23) Event1/Event2 Clean Up ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Use "Local system administrator" in both sections. Discussion: Siva proposed that we clean up the text for these Events by selecting either "Administrator" or "Local system" but not both. Sue proposed text using "Local system administrator" that was agreed on. This was discussed in the "Response to FSM input - Comments 1-10" thread: Comment #4. ---------------------------------------------------------------------------- 24) Events 3, 5, 6 & 7 Give Examples ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Move Automatic Disconnect example to the event description. Possibly add example text/description clarification to other events. Discussion: This began with Siva proposing we add examples for these event states. Sue believes this is largely out-of-scope, but did agree to move the example of "automatic stop" to the event description section. She asked for proposed text for additional examples. This was discussed in the "Response to FSM input - Comments 1-10" thread: Comment #5. ---------------------------------------------------------------------------- 25) Event 4 & 5 Session Initiation Text ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: Leave the text as is. Discussion: This began with Siva wanting to change: Definition: Local system automatically starts the BGP session with the passive flag=20 enabled. The passive flag indicates=20 that the peer will listen prior to=20 establishing a connection. to: The passive flag indicates that the state machine will wait for specified peer to initiate a connection with the local system. If this does not happen within a specific time (hold time), the local system will then also attempt to initiate connection with the specified peer. Sue replied: The text in 8.2.1.1 indicates the definition of the passive flag. 6a) ========== My understanding of your text is that you want to replace in both sets of text: "The passive flag indicates the peer will listen prior to establishing a connection". with: "The passive flag indicates that the state machine will wait for the specified peer to initiate a connection with a local system. The problem with this sentence is that in the "unconfigured" case the phrase "specified" peer is confusing. I think the original text is clearer. 6b) ========== If this does not happen within a specific time (hold time), the local system will then also attempt to initiate (a) connection with the specified peer. My comments: Again, the "specified peer" term is confusing. Also, the 2nd half of the statement mixes the actions of the state machine with the events. I believe this muddies the text instead of clarifying it. Siva and Sue later agreed to leave the text the same because of the Unconfigured + passive TCP connection + Delay Open situation. This was discussed in the "Response to FSM input - Comments 1-10" thread: Comment #6. ---------------------------------------------------------------------------- 26) Event 4 & 5 - bgp_stop_flap option ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Add new event (#3) below. Does that settle things? Discussion: This began with Siva asking: Won't a variant of this with bgp_stop_flap option set be requried ? We can also achieve the same by using the bgp_stop-Flap option as a flag that is provided as an input to the state machine. Siva later clarified this to include: We already have Event 3 - Automatic Start Event 5 - Automatic start with bgp_stop_flap option set To make things consistent, shouldn't we either a) Add 3 new events : 1) Manual start with bgp_stop flap option set 2) Manual start with passive TCP establishment and bgp_stop_flap option set 3) Automatic start with passive TCP establishment and bgp_stop_flap option set or b) Remove Event 6, and rely on a flag to tell us wether peer flap damping is to be performed for the session or not. Sue said she prefered option A. And stated that #1 & #2 are infeasible, but that we need to add #3. This was discussed in the "Response to FSM input - Comments 1-10" thread: Comment #7. ---------------------------------------------------------------------------- 27) Event 5 Clarification ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: Leave the text as is. Discussion: This began when Siva asked that in event 5: Is it correct that this event will occur only when we want to restart a connection (after it had been terminated due to some reason beside administrative action) that we had accepted from an unconfigured peer ? Sue replied: The automatic start function is an implementation specific mechanism. This text does not seek to restrict it in any fashion. Siva said that although he felt his original clarification would be more useful to new implementors he is ok with the text as is. This was discussed in the "Response to FSM input - Comments 1-10" thread: Comment #8. ---------------------------------------------------------------------------- 28) Timer Events Definition - Make Consistant ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Change text to use "generate" across the board. Discussion: Can we use similar language for Events 8-12 to make them consistant? It was agreed that we will use "generate" i.e.: Event 8: An event generated when the Idle Hold timer expires. Event 9: An event generated when the ConnectRetry timer expires. Event 10: An event generated when the Hold timer expires. Event 11: An event generated when the Keepalive timer expires Event 12: An event generated when the Delay BGP Open timer expires. This is at consensus. This was discussed in the "Response to FSM input - Comments 1-10" thread Comment #9. ---------------------------------------------------------------------------- 29) Event 8 - Clean Up ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Clean up first sentance. New text below. Discussion: Siva began this by asking if we could clean up the wording of Event 8. After some discussion with Sue we are at this change for the first sentance: An event trigerred by the expiry of the Idle Hold timer, indicating that the session has completed waiting for a back-off period to prevent bgp peer oscillation. This was discussed in the "Response to FSM input - Comments 1-10" thread: Comment #10. ---------------------------------------------------------------------------- 30) Hold Timer - Split? ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: Keep the hold timer text as is. Discussion: Siva proposed that since: We use the hold timer for two purposes * Waiting for an open message (with a default value of 240 seconds) * Waiting for Keepalives (with a default value of 90 seconds) Can we use two different timers (or at least call them two different timer events) ? Sue replied that this is not how it is implemented currently. Siva replied that we have two conceptually different timers, but that it would certainly work to only have one, since only one needs to be running at any given time. Tom agreed that we can keep things as is. This was discussed in the "Comments 11-20" thread: Comment #11. ---------------------------------------------------------------------------- 31) OpenDelay Timer Definition ---------------------------------------------------------------------------- Status: Consensus Change: Yes - See issue 28 Summary: This is fixed by the fixing of issue 28. Discussion: This began with Siva's request that we add something to Event 12 to specify what to do when the timer expires. This seems to have been addressed in issue 28. This was discussed in the "Comments 11-20" thread: Comment #12. ---------------------------------------------------------------------------- 32) Definition of TCP Connection Accept (Event 13) ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Change "Definition" text as indicated below. Discussion: Siva proposed that we change text from refering to "TCP connection request" to "receiving a TCP connection". This led to this proposed text: Definition: Event indicating the reception of a TCP connection request with a valid source IP address and TCP port, and valid destination IP address and TCP Port. The definition of invalid source address and port and invalid destination address is left to the implementation. This met with agreement. This thread also discussed the idea of filtering the incomming address/port. It was decided that this was implementation dependant. This was discussed in the "Comments 11-20" thread: Comment #13. ---------------------------------------------------------------------------- 33) Event 13 & 14 - Valid Addresses & Ports ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: We're at consensus for most of this, but we need to agree on the destination port portion of this. Discussion: With regard to Event 13 & 14, Siva raised questions about: 1) What does it mean to validate a port, and 2) Should we state what we consider an invalid IP addres to be? Sue replied that this is local policy and is implementation dependant. Siva agreed regarding the source port & IP address, but disagreed about the destination port. He argued that we need to know the destination port for interoperability. This was discussed in the "Comments 11-20" thread: Comment #14. ---------------------------------------------------------------------------- 34) Event 17 - TCP Connection Fails to TCP Connection Termination ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Select either "fails" or "disconnect", and choose title from either "TCP connection fails" or "TCP connection termination". Discussion: This began with Siva observing: This event can occur even when the transport connection is closed by the other end. Since this does not reflect a 'failure ', can we change the event name to % Event17: TCP connection termination Sue replied that: Discussion: It both terminates from the remote site and can "timeout" - fail. Suggestions? I can use "disconnect", what do you think. Siva replied that this was a minor issue, and on further reflection, either "fails" or "disconnect" would be acceptable. This was discussed in the "Comments 11-20" thread: Comment #15. ---------------------------------------------------------------------------- 35) Making Definition Style Consistant ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Almost at consensus with this, just need to decide if we're going to adopt the same style for events 18-21. Text for 13-21 is below Discussion: This started with Siva asking if we could make the definition style consistant across eventsr. Sue replied to this with text for 13-17, Siva clarified that he was talking more about 18-21, and proposed text. We are agreed on the text for 13-17: Event13: TCP connection indication and valid remote peer Definition: Event indicating the local system reception of a TCP connection request with a valid source IP address and TCP port, and valid destination IP address and TCP Port. The definition of invalid source, and invalid destination IP address is left to the implementation. BGP's destination port SHOULD be port 179 as defined by IANA. TCP connection request is denoted by the local system receiving a TCP SYN. Status: Mandatory (Optional) Event14: RCV TCP connection indication with invalid source or destination Definition: Event indicating the local system reception of a TCP connection request with either an invalid source address or port number or an invalid destination address or port number. BGP destination port number SHOULD be 179 as defined by IANA. Again, a TCP connection request denoted by local system receiving a TCP SYN. Status: Mandatory (Optional) Event15: TCP connection request sent received an ACK. Definition: Event indicating the Local system's request to establish a TCP connection to the remote peer. The local system's TCP session sent a TCP SYN, and received a TCP SYN, ACK pair of messages, and Sent a TCP ACK. Status: Mandatory Event16: TCP connection confirmed Definition: Event indicates that the local system receiving a confirmation that the TCP connection has been established by the remote site. The remote peer's TCP engine sent a TCP SYN. The local peer's TCP engine sent a SYN, ACK pair, and now has received a final ACK. Status: Mandatory Event17: TCP connection fails Definition: Event indicates that the local system has received a TCP connection failure notice. The remote BGP peer's TCP machine could have sent a FIN. The local peer would respond with a FIN-ACK. Another alternative is that the local peer indicated a timeout in the TCP session and downed the connection. Status: Mandatory Siva proposed these changes for 18-21: > Event18: BGPOpen > > Definition: An event indicating that a valid Open > message has been received. with % Event18: BGPOpen % % Definition: An event is generated when a valid Open % message has been received. > Event19: BGPOpen with BGP Delay Open Timer running > > Definition: An event indicating that a valid Open > message has been successful > established for a peer that is > currently delaying the sending of an > BGP Open message. with % Event19: BGPOpen with BGP Open Delay Timer running % % Definition: An event is generated when a valid Open % message has been received for a peer that % is currently delaying the sending of a % BGP Open message. Editorial Note: "Delay Open Timer" replaced with "Open Delay Timer" per issue 7. > Event20: BGPHeaderErr > > Definition: BGP message header is not valid. with % Event20: BGPHeaderErr % % Definition: An event is generated when a received BGP % message header is not valid. > Event21: BGPOpenMsgErr > > Definition: An BGP Open message has been received > with errors. with % Event21: BGPOpenMsgErr % % Definition: An event is generated when BGP Open message % with errors has been received. There haven't been any comments back on this text. This was discussed in the "Comments 11-20" thread: Comment #16. ---------------------------------------------------------------------------- 36) Event 19 - Definition Cleanup ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Replace definition for Event 19 with the text in the discussion. Discussion: Siva propsosed we replace: > Definition: An event indicating that a valid Open=20 > Message has been successful=20 > established for a peer that is=20 > currently delaying the sending of an=20 > BGP Open message.=20 with: % Definition: An event indicating that a valid OPEN % Message has been received for a peer % that has a successfully established % transport connection and is currently % delaying the seending of a BGP open % message in Event 19. Sue agreed to the changes. This was discussed in the "Comments 11-20" thread: Comment #17. ---------------------------------------------------------------------------- 37) Event 22 - Cleanup ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Replace Event 22 definition with the text from the discussion. Discussion: Siva began with observing: > Event22: Open collision discard > > Definition: An event generated administratively=20 > when a connection Collision has been=20 > detected while processing an incoming =20 > Open message. This connection has been=20 Isn't this event 'automatically' generated, since it is a system generated event ?=20 Sue replied that: response: How this generated is implementation specific. The "administratively" is to cover policy. Siva also proposed an editorial fix with: > Event 22 is an administrative could=20 > occur if FSM is implemented as two=20 > The word event is missing. How about % Event 22 is an automatic event that % could occur if FSM is implemented as two=20 Sue replied with this rewritten text: Event22: Open collision dump Definition: An event generated administratively when a connection collision has been detected while processing an incoming OPEN message and this connection has been selected to disconnected. See Section 6.8 for more information on collision detection. Event22 is an administrative based only implementation specific policy. This Event may occur if the FSM is implemented as two linked state machines. Sive agreed with this new text. This was discussed in the "Comments 11-20" thread: Comment #18. ---------------------------------------------------------------------------- 38) FSM Description - ConnectRetry Count ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: Leave the counter text alone, since it is used in peer oscillation and will be in the MIB. Discussion: Siva opened with this question: The Connect Retry count is updated by the FSM but never used. In the absence of peer oscillation damping, will this be used to stop connection establishment attempts after a certain maximum number ? <Sue> Yes, this is either implementation specific or is it based on the peer oscillation damping draft. </Sue> Can we include the use of this counter in some place ? <Sue> Connect retry counter 1) Will be utilized by the peer oscillation damping drat. 2) Will be included in bgp-4-mibv2-xx. I just check and I didn't find it. Do you still want text in the main? </Sue> To which Siva replied that he believes we can leave the main text alone. This was discussed in the "Comments 11-20" thread: Comment #19. ---------------------------------------------------------------------------- 39) Handling Event 7 (Auto Stop) to Idle State processing ---------------------------------------------------------------------------- Status: Consensus Change: Yes Summary: Fix the text as indicated in the discussion. Discussion: Siva began with: The handling of Event 7 is missing from the Idle State processing. Can we add this ? How about replacing > An manual stop event (Event2) is ignored in the Idle state. with % Manual stop (Event 2) and Auto stop (Event 7) events are ignored % in the Idle state Sue replied that she would add the text. This was discussed in the "Comments 11-20" thread: Comment #20. ---------------------------------------------------------------------------- 40) Clearing the Connection Retry Timer ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Do we remove the redundant resets from the text? Or leave them to allow action routines to be shared across multiple states? Discussion: Siva opend with the observation: There are a few sections where the FSM draft states that the Connection Retry timer needs to be reset, wheras the conenct retry timer had been cleared prior to entering that state. We can remove these instructions to clear the connect retry timer. List of places where the connect retry timer need not be cleared a) Handling of Event 19 in the Connect State b) Handling of Events 12 in the Active State c) All cases where it is refered to in the OpenSent, OpenConfirm and Established states Sue replied: Comment: 1) Does it hurt to have the connect retry timer cleared at these points, since it has already been cleared. I felt it eased the implementations to allow the action routines to be shared across as many states as possible. You can see this a bit more actively. This was discussed in the "Comments 21-30" thread: Comment #21. ---------------------------------------------------------------------------- 41) Handling of Event 14 in the Connect State ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Is this really part of Issue 13.4? Discussion: Siva opened the discussion with: > If the transport connection receives an indication > that is invalid or unconfigured. [Event 14]: > - the TCP connection is rejected. I don't understand how we would get this event while in this state. Sue replied: See my earlier comments (1-10) on the connection state. It happens in implementations which track the TCP state more closely. I suggest that Event 14 become optional. Sue also suggested we fold this into the discussion about events 13-17, which is tracked in issue 13.4. This was discussed in the "Comments 21-30" thread: Comment #22. ---------------------------------------------------------------------------- 42) Handling events 20, 21 in the Connect State and Active State ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Modify section 4.2 to avoid precluding sending NOTIFICATIONs in early states (text is in the discussion). Still need to clarify the FSM behavior. Discussion: Siva began this with: We need to consider the case where we receive events 20 (message header error) and 21 (Open message error) when the delay timer is running. Since the connection has been established at this point, we need to send a Notification message and then terminate the connection. To which Sue replied: Alternative comments: 1) We have not sent an Open statement. 2) Why do we have to send an Notification? I see no justification for it. Suggestion: Do you have implementations that send notification? Do you know of others that don't. Jeff saw this as indicative of an issue with section 4.2 the way it is currently written: >From section 4.2 of -18: : 4.2 OPEN Message Format : : After a TCP is established, the first message sent by each side is an : OPEN message. If the OPEN message is acceptable, a KEEPALIVE message : confirming the OPEN is sent back. Once the OPEN is confirmed, UPDATE, : KEEPALIVE, and NOTIFICATION messages may be exchanged. This text implies that NOTIFICATIONs can only be sent once we have sent an open and then a keepalive, generally meaning we're in the Established state. Anyone suggestions for modifying the wording? Section 6.1 (Message header error) is one situation that implies that a NOTIFICATION can be sent without sending even an OPEN message. Note that since the base FSM implies that we send an OPEN message immediately when we have a completed trasnport connection, we SHOULD be in at least OpenSent. However, the DelayOpen timer means that we MAY send a NOTIFICATION when we are in the Connect state. GateD, at least, will not send a NOTIFICATION without first sending an OPEN. We need to pick one: You can send NOTIFICATIONS before OPEN or before OPEN if the OpenDelay timer is running. However, we MUST fix the text above. Tom opined: A NOTIFICATION without a preceding OPEN is rather hard to interpret; it is the OPEN that gives the recipient what it needs to know about its potential peer (Version, AS number, ID, options etc) so it makes sense to send an OPEN even if it is followed by a NOTIFICATION to say goodbye :-( as opposed to a KEEPALIVE which says hello:-). But as ever, what is implemented? Yakov suggested these modifications to the text to resolve this: 1. Delete the last sentence in the above paragraph or 2. Delete "and NOTIFICATION" in the last sentence in the above paragraph Jeff replied that he prefered the first option, and that the second could be interpreted as NOTIFICAITONs not being legal, when, in fact, they may. So the text on the table to resolve this is: 4.2 OPEN Message Format After a TCP is established, the first message sent by each side is an OPEN message. If the OPEN message is acceptable, a KEEPALIVE message confirming the OPEN is sent back. However, this does not entirely clear up the original point about the FSM. If we reveive an error in Connect/Active, do we send a NOTIFY? Do we preface it with an OPEN, so that OPEN/NOTIFY are sent in immediete succession? This was discussed in the "Comments 21-30" thread: Comment #23. ---------------------------------------------------------------------------- 43) Handling the default events in the Connect state ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Discussion: Siva opened this with: The Open Delay timer [original: BGP Delay OpenTimers) needs to be cleared if it is running. How about adding this: % - If the ConnectRetry Timer is running % - Clear the Connect Retry timer % - Otherwise % - Clear the Open Delay timer [original: BGP Delay Open Timer] Sue replied that: By the default you mean the text: In response to any other events[Events 7-8, 10-11, 18, 20-27], the local system: "resets" to me implies stops and clears. I think the text is clear than the text above. ------------ Is this the replacement text you imply above: - resets the connect retry timer (sets to zero), - clears the Open Delay timer [original: BGP Delay timer] (sets to zero), - increments the ConnectRetryCnt (connect retry count) by 1, - [optionally] performs bgp peer oscillation damping, and - goes to Idle text: Editor's note: various incarnations of "Open Delay timer" have been replaced with "Open Delay timer". See issue 7. This was discussed in the "Comments 21-30" thread: Comment #24. ---------------------------------------------------------------------------- 44) Handling Event 23 in Connect and OpenSent ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Does the supplied state table suffice to meet this handling of Event 23? Discussion: This began with Siva saying: This is currently being handled in the default event processing section. However, we do not need to go through the peer oscillation damping process in this case. Can we change the wordings to reflect this, or move this out of peer oscillation damping processing ? Sue replied: 1) There is no default event handling process in the text, you will need to specify the text. 2) The state table below (hares-statemt-03.txt) states shows the changes ------------- Event 23 states: current Idle Connect Active Open-Sent Open-Cnf Establish ---------------------------------------------------- next state Idle Idle Idle Idle Idle Idle ----------------------------------------------- action V D D Y Y T ============================================== V - Indicate FSM errors and ignore. D - 1) resets the connect retry timer (sets to zero), 2) drops the TCP connection, 2) releases all BGP resources, 3) increments the ConnectRetryCnt (connect retry count) by 1, 4) [optionally] performs the bgp peer oscillation damping, and Goes to Idle state. Y 1) resets the connect retry timer (sets to zero), 2) Drops the TCP connection, 3) releases all BGP resources, 4) [optionally] This was discussed in the "Comments 21-30" thread: Comment #25. ---------------------------------------------------------------------------- 45) Event 17 in the Connect state ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: We are at consensus, we just need the text. Discussion: This began with Siva asking: > If the transport connection fails (timeout or transport > disconnect) [Event17], the local system: > - changes its state to Active. If the transport connection fails when the Open Delay timer [original: BGP Open Delay timer] is running, should we still be going into the Active state ? Sue replied refering to the disussion tracked in issue 13.4. Jeff responded that: In this particular case, I think the issue is separate from the issues for events 13-17 since this isn't particular to how deep the BGP implementation meddles in the TCP implementation. If we are in the Connect state, because we have an incoming transport connection that has completed, but we have the OpenDelay timer running and the transport connection is closed, we can simply drop into Active after resetting the ConnectRetry timer and clearing the OpenDelay timer (if set/exists). In the case of an unconfigured peer, we can discard the FSM instance. Tom replied that he agreed with this. This was discussed in the "Comments 21-30" thread: Comment #26. ---------------------------------------------------------------------------- 46) Handling of Event 17 in Active state ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: See issue 13.4, this issue closed in favor of that one. Discussion: This began with Siva saying: We should now move into Idle state. Can we add % - Goes to Idle state Sue replied that she thought this should be bundled in with the issue tracked in 13.4. Since no one objected, this issue has been closed in favor of that one. This was discussed in the "Comments 21-30" thread: Comment #27. ---------------------------------------------------------------------------- 47) Handling of Event 19 in Active state ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Which of the texts is prefered? Discussion: This began with Siva suggesting: > - Set the Hold timer to a large value (4 minutes), Since OPEN messages have been exchanged, can we change this to % - If the negotiated Hold time is not 0, set the Hold time to % - the negotiated value Sue replied that: The text in Active and Open Sent needs to be the same. The text in Open Sent is: - sets the Hold timer according to the negotiated value (see section 4.2), and Which text do you prefer? This was discussed in the "Comments 21-30" thread: Comment #28. ---------------------------------------------------------------------------- 48) Handling of Event 2 in Active state ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Clear up procedure for handling Event 2. Discussion: Siva opened with: > A manual stop event[Event2], the local system: > - Sends a notification with a Cease, > - drops the Transport connection These two actions are possible only if a transport connection had already been established. How about changing the text to % - If a transport connection had been successfully established % - Send a Notification with a Cease % - Drop the Transport Connection Sue counter suggested: A manual stop event [Event 2], the local system - Drop the TCP connection, - Release all BGP resources, - resets the connection retry timer [sets to zero], - goes to Idle. Jeff replied: I'm rather confused. Under exactly what circumstances can we be in the Active state and have an active TCP connection at all? Ditto for having any BGP resources? Going to Idle is fine. Tom offered this example: eg start with passive flag, TCP SYN received, SYN-ACK sent, ACK received, Delay Open flag set and there we are. Most events are now possible either from a well-implemented remote peer or a badly implemented remote peer. This was discussed in the "Comments 21-30" thread: Comment #29. ---------------------------------------------------------------------------- 49) Default Event handling in Active state ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: No routes in active. Discussion: Siva began with: To ensure consistencey with E2 handling, can we add % - If any BGP Routes exist, delete the routes Sue replied: Comment: Yakov and Jeff noted, there are no routes in Active state. Since there were no responses disagreeing, we'll consider this closed unless someone wants to open it back up. This was discussed in the "Comments 21-30" thread: Comment #30. ---------------------------------------------------------------------------- 50) Clearing Hold timer in OpenSent, OpenConfirm and Established State ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: This issue is addressed in the "Clear BGP resources" Discussion: This began with Siva stating: In all event handling where we go to Idle state, we need to clear the Hold Timer as well. Sue replied that: issue resolve one way last Jan - March Clearing of keep alive timer included in Clear BGP resources No response to this yet, but since this seems to be resolved it is at consensus unless someone objects. This was discussed in the "Comments 30-36" thread: Comment #31. ---------------------------------------------------------------------------- 51) Clearing Keepalive timer in OpenConfirm and Established State ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: This issue is addressed in the "Clear BGP resources" Discussion: This began with Siva stating: In all event handling where we go to Idle state, we need to clear the Keepalive Timer as well. Sue replied that: issue resolve one way last Jan - March Clearing of keep alive timer included in Clear BGP resources No response to this yet, but since this seems to be resolved it is at consensus unless someone objects. This was discussed in the "Comments 30-36" thread: Comment #32. ---------------------------------------------------------------------------- 52) Handling Event 18 in the OpenSent state (Keepalive Timer) ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: At what point do we start the Keepalive timer. Discussion: This began with Siva asking: Why do we start the Keepalive timer at this stage ? Isn't it sufficient to do so when we move into Established state ? Sue replied: An earlier comment from Tom (and you) requested the following: <---Open [Open sent state] Open--> [Event 18] <---Open <---Keepalive [Action from Event 18 in Open Sent] [Open Confirm] Keepalive -> [Event 25] [established] What do implementations do? We'll have to query implementations. Jeff added: I'm assuming the second OPEN going from right to left is a typo. If it isn't, thats a FSM error to the peer on the left. Theoretically, an implementation that utilizes its keepalive timer to send the first keepalive to transition to Established is still interoperable. However: o Keepalives can be disabled by negotiating hold time of zero o We really shouldn't need to restart the Keepalive timer. If there is a delay in the keepalive that transitions from OpenConfirm to Established, its due to the transport connection. It should be reliable and it *should* get through. If it doesn't, there's other problems and the hold timer for the peer on the right should do the Right Thing and drop the connection. > What do implementations do? We'll have to query implementations. GateD at least waits to enter the Established state prior to starting the KeepAlive timer. Tom also added: My comment was that if we do not send a KeepAlive (and start the KeepAlive timer), on exiting from Active with Event 19 to OpenConfirm then we never will and the connection will die. Open Confirm state means valid Open received so we must send a KeepAlive to acknowledge the Open (as pointed out in Jeff's other posting) and we never do it in OpenConfirm state itself (unless the KeepAlive timer expires which it cannot because we have not started it). So for me, OpenSent state Event 18 was and is correct, sending the KeepAlive without which the connection goes no further and Active state Event 19 needs to be brought into line. To say that the timer is started when entering Established state is fine except for a slight problem; we have no way in this FSM of defining actions that are taken on entering a state, only actions to be taken on leaving another state so that is why the KeepAlive actions need to be where they are (or are not in the case of Active state Event 19). This was discussed in the "Comments 30-36" thread: Comment #33. ---------------------------------------------------------------------------- 53) Established State MIB ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: MIB references pulled in favor of having them in the MIB document. See issue 8. Discussion: This began with Siva asking: Some event handling in the Established state do not set the MIB Reason when handling an event that causes an error. Can we add this ? Sue replied that we have pulled the MIB wording from the FSM. See issue 8. This was discussed in the "Comments 30-36" thread: Comment #34. ---------------------------------------------------------------------------- 54) State impact of not supporting Optional Events ---------------------------------------------------------------------------- Status: No Consensus Change: Potentially Summary: Should we develop the text? If so we need text. Discussion: Siva stated that: For the events whose status is optional, can we state the impact of not supporting them (in terms of any interoperability issues). I understand that most of the optional events will not have such an impact; but a clarification statement for the optional events would benefit new implementors. Sue responded: Much of the support of optional parameters depends on policy. I could put a short note about the optional events and parameters as part of 8.1.5 or 8.2.1.3 I think it fits better in 8.1.5. Optional: Events: 3-8, 12, 13-14[my suggestion] 19, 22 Timers: Idle Hold Timer Open Delay Timer Required flags for optional parameters: Open Delay Flag BGP Stop Flap Sue said she would try to work up more if it is agreed that this is on the right track. This was discussed in the "Comments 30-36" thread: Comment #35. ---------------------------------------------------------------------------- 55) New DelayOpen State ---------------------------------------------------------------------------- Status: Consensus Change: No Summary: We've chosen not to reopen the debate about adding a DelayOpen State to the FSM. Discussion: Siva began with asking: Is delaying the sending of an OPEN message a standrad industry practice ? Also, in the FSM, this has been handled by practically implementing a sub-state each, within the CONNECT and ACTIVE states. Won't the FSM look more simple if we just had a new DelayOpen state that we could move into ? Sue responded that this was something we have tried to do before, but that it spawned some degree of rabid response on both sides. Given our current mandate to stick with what is implemented, it is probably best not to reopen this debate. Unless someone badly wants to reopen this debate, the issue is at Consenus. This was discussed in the "Comments 30-36" thread: Comment #36. --5vNYLRcllDrimb99-- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id TAA05581 for <idr-archive@nic.merit.edu>; Thu, 19 Dec 2002 19:01:35 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id AD61C91232; Thu, 19 Dec 2002 19:00:20 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 7D4619129C; Thu, 19 Dec 2002 19:00:20 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 15B0291232 for <idr@trapdoor.merit.edu>; Thu, 19 Dec 2002 19:00:19 -0500 (EST) Received: by segue.merit.edu (Postfix) id 78DCE5DE6A; Thu, 19 Dec 2002 19:00:18 -0500 (EST) Delivered-To: idr@merit.edu Received: from realname (unknown [203.254.224.24]) by segue.merit.edu (Postfix) with ESMTP id EDA095DDB6 for <idr@merit.edu>; Thu, 19 Dec 2002 19:00:17 -0500 (EST) Received: from custom-daemon.mailout1.samsung.com by mailout1.samsung.com (iPlanet Messaging Server 5.2 HotFix 1.05 (built Nov 6 2002)) id <0H7E00L015C8WJ@mailout1.samsung.com> for idr@merit.edu; Fri, 20 Dec 2002 09:00:08 +0900 (KST) Received: from ep_mmp1 (localhost [127.0.0.1]) by mailout1.samsung.com (iPlanet Messaging Server 5.2 HotFix 1.05 (built Nov 6 2002)) with ESMTP id <0H7E00L515C81C@mailout1.samsung.com> for idr@merit.edu; Fri, 20 Dec 2002 09:00:08 +0900 (KST) Received: from samsungy77z6fd ([165.213.70.137]) by mmp1.samsung.com (iPlanet Messaging Server 5.2 HotFix 1.05 (built Nov 6 2002)) with ESMTPA id <0H7E00G695OBJG@mmp1.samsung.com> for idr@merit.edu; Fri, 20 Dec 2002 09:07:23 +0900 (KST) Date: Fri, 20 Dec 2002 08:59:19 -0800 From: Manav Bhatia <manav@samsung.com> Subject: Re: Error handling in the update message.. To: Shankar Ananthanarayanan Kambat <shankar.kambat@wipro.com> Cc: idr@merit.edu Message-id: <023c01c2a849$23031a00$8946d5a5@samsungy77z6fd> MIME-version: 1.0 X-MIMEOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 X-Mailer: Microsoft Outlook Express 5.00.2919.6700 Content-type: text/plain; charset=iso-8859-1 Content-transfer-encoding: 7BIT X-Priority: 3 X-MSMail-priority: Normal References: <4D148FEC6C003445B94D5B264288DBE949591D@blr-k1-msg.wipro.com> Sender: owner-idr@merit.edu Precedence: bulk Shankar, If a Loc Pref is present in an UPDATE from the external peer then the Loc Pref should be ignored and the rest of the attributes must be processed. We must send a notification only in the extreme cases. --Manav ----- Original Message ----- From: "Shankar Ananthanarayanan Kambat" <shankar.kambat@wipro.com> To: <idr@merit.edu> Sent: Thursday, December 19, 2002 6:04 AM Subject: Error handling in the update message.. Hi, It is well known that LOCAL_PREF and ORIGINATOR_ID(Route Reflector Scenario) should not be sent across autonomous system by BGP. However, if an EBGP speaker receives an update message with such an attribute, may be due to malfunctioning of the peer router, should it ignore the attribute or sent a notification? I don't see an error subcode for the same in the draft/RFC. If it is ignored will it not have any problem? Thanks Shankar K A Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id JAA01849 for <idr-archive@nic.merit.edu>; Thu, 19 Dec 2002 09:05:13 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id B7928912CF; Thu, 19 Dec 2002 09:04:53 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 8CD1D912CE; Thu, 19 Dec 2002 09:04:53 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 45EF2912D1 for <idr@trapdoor.merit.edu>; Thu, 19 Dec 2002 09:04:52 -0500 (EST) Received: by segue.merit.edu (Postfix) id 2072F5DE2C; Thu, 19 Dec 2002 09:04:52 -0500 (EST) Delivered-To: idr@merit.edu Received: from wiprom2mx1.wipro.com (wiprom2mx1.wipro.com [203.197.164.41]) by segue.merit.edu (Postfix) with ESMTP id BD1FF5DDAF for <idr@merit.edu>; Thu, 19 Dec 2002 09:04:48 -0500 (EST) Received: from m2vwall2 (m2vwall2.wipro.com [164.164.29.236]) by wiprom2mx1.wipro.com (8.11.3/8.11.3) with SMTP id gBJE4ic14651 for <idr@merit.edu>; Thu, 19 Dec 2002 19:34:44 +0530 (IST) Received: from blr-k1-msg.wipro.com ([10.117.50.99]) by blr-m1-bh1.wipro.com with Microsoft SMTPSVC(5.0.2195.4905); Thu, 19 Dec 2002 19:34:42 +0530 content-class: urn:content-classes:message MIME-Version: 1.0 Content-Type: multipart/mixed; boundary="----=_NextPartTM-000-81b2487a-9120-4fd9-99f7-af568ea83ca9" Subject: Error handling in the update message.. X-MimeOLE: Produced By Microsoft Exchange V6.0.5762.3 Date: Thu, 19 Dec 2002 19:34:42 +0530 Message-ID: <4D148FEC6C003445B94D5B264288DBE949591D@blr-k1-msg.wipro.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: Error handling in the update message.. Thread-Index: AcKmq4jlO2oyIhOZSrS2ldGxuV73NAAuvriQ From: "Shankar Ananthanarayanan Kambat" <shankar.kambat@wipro.com> To: <idr@merit.edu> X-OriginalArrivalTime: 19 Dec 2002 14:04:42.0636 (UTC) FILETIME=[940104C0:01C2A767] Sender: owner-idr@merit.edu Precedence: bulk This is a multi-part message in MIME format. ------=_NextPartTM-000-81b2487a-9120-4fd9-99f7-af568ea83ca9 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi, It is well known that LOCAL_PREF and ORIGINATOR_ID(Route Reflector = Scenario) should not be sent across autonomous system by BGP. However, = if an EBGP speaker receives an update message with such an attribute, = may be due to malfunctioning of the peer router, should it ignore the = attribute or sent a notification? I don't see an error subcode for the = same in the draft/RFC. If it is ignored will it not have any problem? Thanks=20 Shankar K A ------=_NextPartTM-000-81b2487a-9120-4fd9-99f7-af568ea83ca9 Content-Type: text/plain; name="Wipro_Disclaimer.txt" Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename="Wipro_Disclaimer.txt" **************************Disclaimer************************************************** Information contained in this E-MAIL being proprietary to Wipro Limited is 'privileged' and 'confidential' and intended for use only by the individual or entity to which it is addressed. You are notified that any use, copying or dissemination of the information contained in the E-MAIL in any manner whatsoever is strictly prohibited. **************************************************************************************** ------=_NextPartTM-000-81b2487a-9120-4fd9-99f7-af568ea83ca9-- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id KAA17042 for <idr-archive@nic.merit.edu>; Wed, 18 Dec 2002 10:38:28 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id DF4C5912AF; Wed, 18 Dec 2002 10:38:04 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id A8C82912B0; Wed, 18 Dec 2002 10:38:04 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 3B649912AF for <idr@trapdoor.merit.edu>; Wed, 18 Dec 2002 10:38:03 -0500 (EST) Received: by segue.merit.edu (Postfix) id 1E2BB5DF84; Wed, 18 Dec 2002 10:38:03 -0500 (EST) Delivered-To: idr@merit.edu Received: from mailscan.telica.com (bouncer.telica.com [4.19.224.197]) by segue.merit.edu (Postfix) with ESMTP id 7A7CA5DFD3 for <idr@merit.edu>; Wed, 18 Dec 2002 10:38:02 -0500 (EST) Received: (from root@localhost) by mailscan.telica.com (8.11.6/8.11.6) id gBIFTPg07617; Wed, 18 Dec 2002 10:29:25 -0500 Received: from ran.ietf.org (ran.ietf.org [132.151.6.60]) by mailscan.telica.com (8.11.6/8.11.6) with SMTP id gBIEahV04972 for <wkwan@telica.com>; Wed, 18 Dec 2002 09:36:43 -0500 Received: from majordomo by ran.ietf.org with local (Exim 4.10) id 18Oe3t-0006fv-00 for ietf-announce-list@ran.ietf.org; Wed, 18 Dec 2002 08:16:33 -0500 Received: from odin.ietf.org ([10.27.2.28] helo=ietf.org) by ran.ietf.org with esmtp (Exim 4.10) id 18Odvn-0006Qq-00 for all-ietf@ran.ietf.org; Wed, 18 Dec 2002 08:08:11 -0500 Received: from CNRI.Reston.VA.US (localhost [127.0.0.1]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA04734; Wed, 18 Dec 2002 08:04:44 -0500 (EST) Message-Id: <200212181304.IAA04734@ietf.org> Mime-Version: 1.0 Content-Type: Multipart/Mixed; Boundary="NextPart" To: IETF-Announce: ; Cc: idr@merit.edu From: Internet-Drafts@ietf.org Reply-To: Internet-Drafts@ietf.org Subject: I-D ACTION:draft-ietf-idr-as4bytes-06.txt Date: Wed, 18 Dec 2002 08:04:43 -0500 Sender: owner-idr@merit.edu Precedence: bulk --NextPart A New Internet-Draft is available from the on-line Internet-Drafts directories. This draft is a work item of the Inter-Domain Routing Working Group of the IETF. Title : BGP support for four-octet AS number space Author(s) : Q. Vohra, E. Chen Filename : draft-ietf-idr-as4bytes-06.txt Pages : 7 Date : 2002-12-16 Currently the Autonomous System number is encoded in BGP [BGP] as a two-octets field. This document describes extensions to BGP to carry the Autonomous System number as a four-octets field. A URL for this Internet-Draft is: http://www.ietf.org/internet-drafts/draft-ietf-idr-as4bytes-06.txt To remove yourself from the IETF Announcement list, send a message to ietf-announce-request with the word unsubscribe in the body of the message. Internet-Drafts are also available by anonymous FTP. Login with the username "anonymous" and a password of your e-mail address. After logging in, type "cd internet-drafts" and then "get draft-ietf-idr-as4bytes-06.txt". A list of Internet-Drafts directories can be found in http://www.ietf.org/shadow.html or ftp://ftp.ietf.org/ietf/1shadow-sites.txt Internet-Drafts can also be obtained by e-mail. Send a message to: mailserv@ietf.org. In the body type: "FILE /internet-drafts/draft-ietf-idr-as4bytes-06.txt". NOTE: The mail server at ietf.org can return the document in MIME-encoded form by using the "mpack" utility. To use this feature, insert the command "ENCODING mime" before the "FILE" command. To decode the response(s), you will need "munpack" or a MIME-compliant mail reader. Different MIME-compliant mail readers exhibit different behavior, especially when dealing with "multipart" MIME messages (i.e. documents which have been split up into multiple messages), so check your local documentation on how to manipulate these messages. Below is the data which will enable a MIME compliant mail reader implementation to automatically retrieve the ASCII version of the Internet-Draft. --NextPart Content-Type: Multipart/Alternative; Boundary="OtherAccess" --OtherAccess Content-Type: Message/External-body; access-type="mail-server"; server="mailserv@ietf.org" Content-Type: text/plain Content-ID: <2002-12-16143303.I-D@ietf.org> ENCODING mime FILE /internet-drafts/draft-ietf-idr-as4bytes-06.txt --OtherAccess Content-Type: Message/External-body; name="draft-ietf-idr-as4bytes-06.txt"; site="ftp.ietf.org"; access-type="anon-ftp"; directory="internet-drafts" Content-Type: text/plain Content-ID: <2002-12-16143303.I-D@ietf.org> --OtherAccess-- --NextPart-- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id IAA09187 for <idr-archive@nic.merit.edu>; Wed, 18 Dec 2002 08:08:22 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 6474691226; Wed, 18 Dec 2002 08:07:48 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 34627912A9; Wed, 18 Dec 2002 08:07:48 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 0C29191226 for <idr@trapdoor.merit.edu>; Wed, 18 Dec 2002 08:07:46 -0500 (EST) Received: by segue.merit.edu (Postfix) id EDAB15DFCD; Wed, 18 Dec 2002 08:07:45 -0500 (EST) Delivered-To: idr@merit.edu Received: from ietf.org (odin.ietf.org [132.151.1.176]) by segue.merit.edu (Postfix) with ESMTP id 74A765DEF1 for <idr@merit.edu>; Wed, 18 Dec 2002 08:07:45 -0500 (EST) Received: from CNRI.Reston.VA.US (localhost [127.0.0.1]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA04734; Wed, 18 Dec 2002 08:04:44 -0500 (EST) Message-Id: <200212181304.IAA04734@ietf.org> Mime-Version: 1.0 Content-Type: Multipart/Mixed; Boundary="NextPart" To: IETF-Announce: ; Cc: idr@merit.edu From: Internet-Drafts@ietf.org Reply-To: Internet-Drafts@ietf.org Subject: I-D ACTION:draft-ietf-idr-as4bytes-06.txt Date: Wed, 18 Dec 2002 08:04:43 -0500 Sender: owner-idr@merit.edu Precedence: bulk --NextPart A New Internet-Draft is available from the on-line Internet-Drafts directories. This draft is a work item of the Inter-Domain Routing Working Group of the IETF. Title : BGP support for four-octet AS number space Author(s) : Q. Vohra, E. Chen Filename : draft-ietf-idr-as4bytes-06.txt Pages : 7 Date : 2002-12-16 Currently the Autonomous System number is encoded in BGP [BGP] as a two-octets field. This document describes extensions to BGP to carry the Autonomous System number as a four-octets field. A URL for this Internet-Draft is: http://www.ietf.org/internet-drafts/draft-ietf-idr-as4bytes-06.txt To remove yourself from the IETF Announcement list, send a message to ietf-announce-request with the word unsubscribe in the body of the message. Internet-Drafts are also available by anonymous FTP. Login with the username "anonymous" and a password of your e-mail address. After logging in, type "cd internet-drafts" and then "get draft-ietf-idr-as4bytes-06.txt". A list of Internet-Drafts directories can be found in http://www.ietf.org/shadow.html or ftp://ftp.ietf.org/ietf/1shadow-sites.txt Internet-Drafts can also be obtained by e-mail. Send a message to: mailserv@ietf.org. In the body type: "FILE /internet-drafts/draft-ietf-idr-as4bytes-06.txt". NOTE: The mail server at ietf.org can return the document in MIME-encoded form by using the "mpack" utility. To use this feature, insert the command "ENCODING mime" before the "FILE" command. To decode the response(s), you will need "munpack" or a MIME-compliant mail reader. Different MIME-compliant mail readers exhibit different behavior, especially when dealing with "multipart" MIME messages (i.e. documents which have been split up into multiple messages), so check your local documentation on how to manipulate these messages. Below is the data which will enable a MIME compliant mail reader implementation to automatically retrieve the ASCII version of the Internet-Draft. --NextPart Content-Type: Multipart/Alternative; Boundary="OtherAccess" --OtherAccess Content-Type: Message/External-body; access-type="mail-server"; server="mailserv@ietf.org" Content-Type: text/plain Content-ID: <2002-12-16143303.I-D@ietf.org> ENCODING mime FILE /internet-drafts/draft-ietf-idr-as4bytes-06.txt --OtherAccess Content-Type: Message/External-body; name="draft-ietf-idr-as4bytes-06.txt"; site="ftp.ietf.org"; access-type="anon-ftp"; directory="internet-drafts" Content-Type: text/plain Content-ID: <2002-12-16143303.I-D@ietf.org> --OtherAccess-- --NextPart-- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id LAA06613 for <idr-archive@nic.merit.edu>; Mon, 16 Dec 2002 11:07:03 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id D274A9123D; Mon, 16 Dec 2002 11:06:25 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id A408C91261; Mon, 16 Dec 2002 11:06:25 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 2C4159123D for <idr@trapdoor.merit.edu>; Mon, 16 Dec 2002 11:06:24 -0500 (EST) Received: by segue.merit.edu (Postfix) id 103A85DF3D; Mon, 16 Dec 2002 11:06:24 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id AA65D5DE85 for <idr@merit.edu>; Mon, 16 Dec 2002 11:06:23 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gBGG6MT19799; Mon, 16 Dec 2002 11:06:22 -0500 (EST) (envelope-from jhaas@jhaas.nexthop.com) Received: from jhaas.nexthop.com (jhaas.nexthop.com [64.211.218.31]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gBGG6IC19792; Mon, 16 Dec 2002 11:06:18 -0500 (EST) (envelope-from jhaas@jhaas.nexthop.com) Received: (from jhaas@localhost) by jhaas.nexthop.com (8.11.3nb1/8.11.3) id gBGG6Ik03488; Mon, 16 Dec 2002 11:06:18 -0500 (EST) Date: Mon, 16 Dec 2002 11:06:18 -0500 From: Jeffrey Haas <jhaas@nexthop.com> To: Yakov Rekhter <yakov@juniper.net> Cc: idr@merit.edu Subject: Re: Comments 21-30 Message-ID: <20021216110617.A26056@nexthop.com> References: <20021212180748.Q6330@nexthop.com> <200212161603.gBGG3TS85345@merlot.juniper.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <200212161603.gBGG3TS85345@merlot.juniper.net>; from yakov@juniper.net on Mon, Dec 16, 2002 at 08:03:29AM -0800 X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Yakov, On Mon, Dec 16, 2002 at 08:03:29AM -0800, Yakov Rekhter wrote: > > From section 4.2 of -18: > > : 4.2 OPEN Message Format > > : > > : After a TCP is established, the first message sent by each side is an > > : OPEN message. If the OPEN message is acceptable, a KEEPALIVE message > > : confirming the OPEN is sent back. Once the OPEN is confirmed, UPDATE, > > : KEEPALIVE, and NOTIFICATION messages may be exchanged. > > > > This text implies that NOTIFICATIONs can only be sent once we > > have sent an open and then a keepalive, generally meaning we're in the > > Established state. > > > > Anyone suggestions for modifying the wording? > > I have two suggestions: > > 1. Delete the last sentence in the above paragraph I would suggest this option. > 2. Delete "and NOTIFICATION" in the last sentence in the above paragraph This option could be interpreted as NOTIFICATIONs not being legal here. -- Jeff -- Jeff Haas NextHop Technologies Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id LAA06443 for <idr-archive@nic.merit.edu>; Mon, 16 Dec 2002 11:03:53 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 3872E9122D; Mon, 16 Dec 2002 11:03:33 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 042199123D; Mon, 16 Dec 2002 11:03:32 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id CBB569122D for <idr@trapdoor.merit.edu>; Mon, 16 Dec 2002 11:03:31 -0500 (EST) Received: by segue.merit.edu (Postfix) id B9AE35DE85; Mon, 16 Dec 2002 11:03:31 -0500 (EST) Delivered-To: idr@merit.edu Received: from merlot.juniper.net (natint.juniper.net [207.17.136.129]) by segue.merit.edu (Postfix) with ESMTP id 620BF5DDA8 for <idr@merit.edu>; Mon, 16 Dec 2002 11:03:31 -0500 (EST) Received: from juniper.net (garnet.juniper.net [172.17.28.17]) by merlot.juniper.net (8.11.3/8.11.3) with ESMTP id gBGG3TS85345; Mon, 16 Dec 2002 08:03:29 -0800 (PST) (envelope-from yakov@juniper.net) Message-Id: <200212161603.gBGG3TS85345@merlot.juniper.net> To: Jeffrey Haas <jhaas@nexthop.com> Cc: Susan Hares <shares@nexthop.com>, idr@merit.edu Subject: Re: Comments 21-30 In-Reply-To: Your message of "Thu, 12 Dec 2002 18:07:48 EST." <20021212180748.Q6330@nexthop.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <50605.1040054609.1@juniper.net> Date: Mon, 16 Dec 2002 08:03:29 -0800 From: Yakov Rekhter <yakov@juniper.net> Sender: owner-idr@merit.edu Precedence: bulk Jeff, > > 23) Handling events 20, 21 in the Connect State and Active State > > > > We need to consider the case where we receive events 20 (message > > header error) and 21 (Open message error) when the delay timer is > > running. > > > > Since the connection has been established at this point, we need to > > send a Notification message and then terminate the connection. > > > > Alternative comments: > > > > 1) We have not sent an Open statement. > > 2) Why do we have to send an Notification? I see no justification > > for it. > > > > Suggestion: > > Do you have implementations that send notification? Do > > you know of others that don't. > > This points out a potential issue. > From section 4.2 of -18: > : 4.2 OPEN Message Format > : > : After a TCP is established, the first message sent by each side is an > : OPEN message. If the OPEN message is acceptable, a KEEPALIVE message > : confirming the OPEN is sent back. Once the OPEN is confirmed, UPDATE, > : KEEPALIVE, and NOTIFICATION messages may be exchanged. > > This text implies that NOTIFICATIONs can only be sent once we > have sent an open and then a keepalive, generally meaning we're in the > Established state. > > Anyone suggestions for modifying the wording? I have two suggestions: 1. Delete the last sentence in the above paragraph or 2. Delete "and NOTIFICATION" in the last sentence in the above paragraph Yakov. > Section 6.1 (Message header error) is one situation that implies > that a NOTIFICATION can be sent without sending even an OPEN message. > Note that since the base FSM implies that we send an OPEN message > immediately when we have a completed trasnport connection, we SHOULD > be in at least OpenSent. However, the DelayOpen timer means that we > MAY send a NOTIFICATION when we are in the Connect state. > > GateD, at least, will not send a NOTIFICATION without first sending > an OPEN. > > We need to pick one: You can send NOTIFICATIONS before OPEN or before > OPEN if the OpenDelay timer is running. However, we MUST fix the text > above. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id RAA15836 for <idr-archive@nic.merit.edu>; Fri, 13 Dec 2002 17:49:53 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 03546912FD; Fri, 13 Dec 2002 17:49:12 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id AF4A891240; Fri, 13 Dec 2002 17:49:11 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id A279391214 for <idr@trapdoor.merit.edu>; Fri, 13 Dec 2002 17:47:39 -0500 (EST) Received: by segue.merit.edu (Postfix) id 7F5A75DEBD; Fri, 13 Dec 2002 17:47:39 -0500 (EST) Delivered-To: idr@merit.edu Received: from nemesis.systems.pipex.net (nemesis.systems.pipex.net [62.241.160.8]) by segue.merit.edu (Postfix) with ESMTP id 32E0A5DE5A for <idr@merit.edu>; Fri, 13 Dec 2002 17:47:39 -0500 (EST) Received: from tom3 (useran15.uk.uudial.com [62.188.135.32]) by nemesis.systems.pipex.net (Postfix) with SMTP id 7240616007F8F; Fri, 13 Dec 2002 22:47:32 +0000 (GMT) Message-ID: <011301c2a2f9$94e53180$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: "Jeffrey Haas" <jhaas@nexthop.com>, "Susan Hares" <shares@nexthop.com> Cc: <idr@merit.edu> Subject: Re: Comments 21-30 Date: Fri, 13 Dec 2002 22:29:23 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Comments in-line on 23 and 29 Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Jeffrey Haas <jhaas@nexthop.com> To: Susan Hares <shares@nexthop.com> Cc: idr@merit.edu <idr@merit.edu> Date: 12 December 2002 23:08 Subject: Re: Comments 21-30 >On Thu, Dec 05, 2002 at 04:21:00PM -0500, Susan Hares wrote: >> 23) Handling events 20, 21 in the Connect State and Active State >> >> We need to consider the case where we receive events 20 (message >> header error) and 21 (Open message error) when the delay timer is >> running. >> >> Since the connection has been established at this point, we need to >> send a Notification message and then terminate the connection. >> >> Alternative comments: >> >> 1) We have not sent an Open statement. >> 2) Why do we have to send an Notification? I see no justification >> for it. >> >> Suggestion: >> Do you have implementations that send notification? Do >> you know of others that don't. > >This points out a potential issue. >From section 4.2 of -18: >: 4.2 OPEN Message Format >: >: After a TCP is established, the first message sent by each side is an >: OPEN message. If the OPEN message is acceptable, a KEEPALIVE message >: confirming the OPEN is sent back. Once the OPEN is confirmed, UPDATE, >: KEEPALIVE, and NOTIFICATION messages may be exchanged. > >This text implies that NOTIFICATIONs can only be sent once we >have sent an open and then a keepalive, generally meaning we're in the >Established state. > >Anyone suggestions for modifying the wording? > >Section 6.1 (Message header error) is one situation that implies >that a NOTIFICATION can be sent without sending even an OPEN message. >Note that since the base FSM implies that we send an OPEN message >immediately when we have a completed trasnport connection, we SHOULD >be in at least OpenSent. However, the DelayOpen timer means that we >MAY send a NOTIFICATION when we are in the Connect state. > >GateD, at least, will not send a NOTIFICATION without first sending >an OPEN. > >We need to pick one: You can send NOTIFICATIONS before OPEN or before >OPEN if the OpenDelay timer is running. However, we MUST fix the text >above. A NOTIFICATION without a preceding OPEN is rather hard to interpret; it is the OPEN that gives the recipient what it needs to know about its potential peer (Version, AS number, ID, options etc) so it makes sense to send an OPEN even if it is followed by a NOTIFICATION to say goodbye :-( as opposed to a KEEPALIVE which says hello:-). But as ever, what is implemented? > >> ============================= >> >> 26) Handling Event 17 in The connect state >> >> > If the transport connection fails (timeout or transport >> > disconnect) [Event17], the local system: >> > - changes its state to Active.=20 >> >> >If the transport connection fails when the BGP Open Delay timer is >> > running, should we still be going into the Active state ? >> >> see mail thread on eventse 13-17 > >In this particular case, I think the issue is separate from the issues >for events 13-17 since this isn't particular to how deep the BGP >implementation meddles in the TCP implementation. > >If we are in the Connect state, because we have an incoming transport >connection that has completed, but we have the OpenDelay timer >running and the transport connection is closed, we can simply >drop into Active after resetting the ConnectRetry timer and clearing >the OpenDelay timer (if set/exists). In the case of an unconfigured >peer, we can discard the FSM instance. > Yup, makes sense to me >> ========================================= >> >> 29) Handling Event 2 in Active state >> >> >> A manual stop event[Event2], the local system: >> >> - Sends a notification with a Cease, >> >> - drops the Transport connection,=20 >> > >> >These two actions are possible only if a transport connection had = >> > already been established. How about changing the text to >> > >> > % - If a transport connection had been successfully established >> > % - Send a Notification with a Cease >> > % - Drop the Transport Connection >> >> How about this text: >> >> A manual stop event [Event 2], the local system >> - Drop the TCP connection, >> - Release all BGP resources, >> - resets the connection retry timer [sets to zero], >> - goes to Idle. > >I'm rather confused. Under exactly what circumstances can we be >in the Active state and have an active TCP connection at all? >Ditto for having any BGP resources? > >Going to Idle is fine. > > eg start with passive flag, TCP SYN received, SYN-ACK sent, ACK received, Delay Open flag set and there we are. Most events are now possible either from a well-implemented remote peer or a badly implemented remote peer. >-- >Jeff Haas >NextHop Technologies Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id RAA15810 for <idr-archive@nic.merit.edu>; Fri, 13 Dec 2002 17:49:29 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id C723191214; Fri, 13 Dec 2002 17:49:11 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 37769912E1; Fri, 13 Dec 2002 17:49:11 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 507A491240 for <idr@trapdoor.merit.edu>; Fri, 13 Dec 2002 17:47:42 -0500 (EST) Received: by segue.merit.edu (Postfix) id 341F25DEBD; Fri, 13 Dec 2002 17:47:42 -0500 (EST) Delivered-To: idr@merit.edu Received: from nemesis.systems.pipex.net (nemesis.systems.pipex.net [62.241.160.8]) by segue.merit.edu (Postfix) with ESMTP id DDD0F5DE5A for <idr@merit.edu>; Fri, 13 Dec 2002 17:47:41 -0500 (EST) Received: from tom3 (useran15.uk.uudial.com [62.188.135.32]) by nemesis.systems.pipex.net (Postfix) with SMTP id 76506160080B2; Fri, 13 Dec 2002 22:47:35 +0000 (GMT) Message-ID: <011401c2a2f9$96028860$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: "Jeffrey Haas" <jhaas@nexthop.com>, "Susan Hares" <shares@nexthop.com> Cc: <idr@merit.edu> Subject: Re: Comments 30-36 from Siva Date: Fri, 13 Dec 2002 22:46:06 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Sue, Jeff My comment was that if we do not send a KeepAlive (and start the KeepAlive timer), on exiting from Active with Event 19 to OpenConfirm then we never will and the connection will die. Open Confirm state means valid Open received so we must send a KeepAlive to acknowledge the Open (as pointed out in Jeff's other posting) and we never do it in OpenConfirm state itself (unless the KeepAlive timer expires which it cannot because we have not started it). So for me, OpenSent state Event 18 was and is correct, sending the KeepAlive without which the connection goes no further and Active state Event 19 needs to be brought into line. To say that the timer is started when entering Established state is fine except for a slight problem; we have no way in this FSM of defining actions that are taken on entering a state, only actions to be taken on leaving another state so that is why the KeepAlive actions need to be where they are (or are not in the case of Active state Event 19). Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Jeffrey Haas <jhaas@nexthop.com> To: Susan Hares <shares@nexthop.com> Cc: idr@merit.edu <idr@merit.edu> Date: 12 December 2002 23:22 Subject: Re: Comments 30-36 from Siva >Sue, > >On Thu, Dec 05, 2002 at 04:18:18PM -0500, Susan Hares wrote: >> 33) Handling Event 18 in the OpenSent state >> >> >Why do we start the Keepalive timer at this stage ? Isn't it sufficient = >> >to do so when we move into Established state ? >> >> An earlier comment from Tom (and you) requested the following: >> >> <---Open >> [Open sent state] >> >> Open--> >> [Event 18] >> >> <---Open >> <---Keepalive >> [Action from Event 18 in Open Sent] >> [Open Confirm] >> Keepalive -> [Event 25] >> >> [established] > >I'm assuming the second OPEN going from right to left is a typo. >If it isn't, thats a FSM error to the peer on the left. > >Theoretically, an implementation that utilizes its keepalive timer >to send the first keepalive to transition to Established is >still interoperable. However: > >o Keepalives can be disabled by negotiating hold time of zero >o We really shouldn't need to restart the Keepalive timer. > If there is a delay in the keepalive that transitions from > OpenConfirm to Established, its due to the transport connection. > It should be reliable and it *should* get through. If it > doesn't, there's other problems and the hold timer for the > peer on the right should do the Right Thing and drop the > connection. > >> What do implementations do? We'll have to query implementations. > >GateD at least waits to enter the Established state prior to starting >the KeepAlive timer. > >-- >Jeff Haas >NextHop Technologies Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id QAA12792 for <idr-archive@nic.merit.edu>; Fri, 13 Dec 2002 16:52:22 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 5787C912FC; Fri, 13 Dec 2002 16:51:46 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 249F2912FD; Fri, 13 Dec 2002 16:51:46 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 05554912FC for <idr@trapdoor.merit.edu>; Fri, 13 Dec 2002 16:51:36 -0500 (EST) Received: by segue.merit.edu (Postfix) id E36A35E005; Fri, 13 Dec 2002 16:51:35 -0500 (EST) Delivered-To: idr@merit.edu Received: from nemesis.systems.pipex.net (nemesis.systems.pipex.net [62.241.160.8]) by segue.merit.edu (Postfix) with ESMTP id 332785DDFF for <idr@merit.edu>; Fri, 13 Dec 2002 16:51:35 -0500 (EST) Received: from tom3 (userbf91.uk.uudial.com [62.188.142.112]) by nemesis.systems.pipex.net (Postfix) with SMTP id CA27016007448; Fri, 13 Dec 2002 21:51:24 +0000 (GMT) Message-ID: <004c01c2a2f1$be159a20$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: "Susan Hares" <shares@nexthop.com>, "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com>, "Yakov Rekhter" <yakov@juniper.net>, "Jeffrey Haas" <jhaas@nexthop.com> Cc: "idr" <idr@merit.edu>, "Venu Kumar G - CTD, Chennai." <venug@ctd.hcltech.com> Subject: Re: Response to FSM input - Comments 1-10 Date: Fri, 13 Dec 2002 21:49:56 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Sue, Siva Please can we stay with the terms Open Delay timer peer oscillation damping in any new text (as previously agreed on the list). There are some new variations confusing their way into the text you are suggesting below. Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Susan Hares <shares@nexthop.com> To: Sivananda Ramnath - CTD, Chennai. <siva@ctd.hcltech.com>; Yakov Rekhter <yakov@juniper.net>; Jeffrey Haas <jhaas@nexthop.com> Cc: idr@merit.edu <idr@merit.edu>; Venu Kumar G - CTD, Chennai. <venug@ctd.hcltech.com> Date: 09 December 2002 21:46 Subject: RE: Response to FSM input - Comments 1-10 Siva: Thanks for such a quick turn around. So.. our summary is 1 - consensus, no text change 2 - no consensus, see text change 3 - mostly consensus, Policy in implementations may dictate some options per peer and some options per implementation. I suggest we specify the list you came up with plus Idle Hold Timer in 8.0. 4 - consensus, text change [done] 5 - no consensus, see comment below 5a - answer question [no text intended] 5b.1 - text proposed 5b.2 - your text or mine? 6 - consensus on text, answer given below 7 - no consensus, text proposal below 2 of the 3 events could be considered infeasible. The 3rd event all add. Please review the text. 8 - consensus, no text change 9 - consensus, text change [done] 10 - consensus, text changed [done] Note 9 + 10 are taken together Comments to resolve: 2,3, 5, 7, Thanks for the quick response. Sue Hares ===================== Comment 2 - Event8: Idle hold timer expires Definition: An Event generated when the Idle Hold Timer expires. The Idle Hold Timer is only used when the persistent peer oscillation damping function is enabled. % % Implementations not implemented persistent % peer oscillations damping functions may not % have the Idle Hold Timer. Is this approach ok? ============ Comment 3 The following will be added to 8.0 Optional parameters that may be supported either per connection or per implementation: 1) Delay Open flag 2) Delay Open Timer 3) Perform automatic start flag 4) Passive TCP establishment flag 5) BGP stop_peer_flag flag %6) Idle Hold timer 7) Perform automatic stop flag 8) Perform Collision detect in Establish mode flag ===================== Comment 4: consensus, change made to Event 1 and 2 =============== Comment 5: 5a: Automatic stop and stop are implemented by cisco's policy engine. The origination of the damping peer oscillation was in result to a bug that was listed in comment 5. 5b: 1.) Example of "automatic disconnect" in Established state Is the text you wanted moved: An example automatic stop event is exceeding the number of prefixes for a given peer and the local system automatically disconnecting the peer. If so, I can move this to the automatic stop event 2.) Text describing the optional parameters. We can add text for the optional parameters to describe their function. Can you propose some text? ====================== Comment 6 > I am ok to leave the text as is, based on text in 8.2.1.1. However, > I do not quite understand your point. In the case of an un-configured > peer, the local system cannot 'establish a connection', so is the > unconfigured peer case still valid in this context ? Unconfigured + passive TCP connection + Delay Open is valid. ============================================ Comment 7 > > > We already have > Event 3 - Automatic Start > Event 5 - Automatic start with bgp_stop_flap option set > > > To make things consistent, shouldn't we either > > a) Add 3 new events : 1) Manual start with bgp_stop flap option set > 2) Manual start with passive TCP establishment > and bgp_stop_flap option set > 3) Automatic start with passive TCP establishment > and bgp_stop_flap option set I like option A, it is clearer. The Manual start or stop terminates the bgp peer oscillation damping. The theory is the operator "did" something, therefore the events need to chagne. Since these events do not have valid implementation in the peer flap damping, I did not include them. #3 is something I missed. I will add that event. Is that resolution OK, or I can put all 3 new events in. ========================= -----Original Message----- From: Sivananda Ramnath - CTD, Chennai. [mailto:siva@ctd.hcltech.com] Sent: Friday, December 06, 2002 6:05 AM To: Susan Hares; Yakov Rekhter; Jeffrey Haas Cc: idr@merit.edu; Sivananda Ramnath (E-mail); Venu Kumar G - CTD, Chennai. Subject: RE: Response to FSM input - Comments 1-10 Hello, Revised summary: Comment 1 - No text change, Potential consensus Comment 2a - No consensus Comment 2b - No consensus Comment 3 - New text suggested, no consensus Comment 4 - Resolved, text changed. Can we consider this closed ? Comment 5 - No text change, Potential consensus Comment 6 - No consensus Comment 7 - No consensus, comment re-phrased. Comment 8 - No text change, potential consensus Comment 9 - Resolved, text changed. Can we consider this closed ? Comment 10 - Resolved, text changed. Can we consider this closed ? > > Comment 1 - no consensus > Comment 2a - out of date (No IdleHold Timer exists) > Comment 2b - no consensus, can't remove due to OpenSent state > action. > > Comment 3 - no text proposed, do we want all options > specified in 8.1 (7 options) > Please review and comment on text > > Comment 4 - agree > Comment 5 - no text proposed, no consensus > Comment 6 - no consensus on the text > Comment 7 - see comment 3, no consensus > Comment 8 - question, no text > Comment 9 - agree, I supplied text > Comment 10 - agree to remove text in option, > different text for definition. > > > Sue Hares > > ========== > > comment 1: > > 1) removal of peer oscillation damping > > The reason we should not remove the peer oscillation damping > from the state machine: > > 1) Deployed implementations support peer oscillation damping > 2) Hooks for the additions in the FSM cannot be added later. > > These hooks are optional and do not need to be implemented. I understand. I am not trying to object to peer oscillation damping, I think it is a good idea and we have included it in our implementation as well. I was suggesting that instead of a partial description in this draft, it be completely described in the draft on peer oscillation damping. However, I do see your point, and unless there are any objections from others, I think we have consensus on this issue. > > > 2) Idle HoldTimer in draft-18 > > Perhaps this is an old comment. I don't find IdleHoldTimer > in the messages. It is listed as Event 8. Perhaps this is what needs to be removed ? > > 3) HoldTime and HoldTimer > > The Hold Time is specified on page 13 of the draft-18 specification > as to the value and size. The text in OpenSent requires > both a HoldTime and HoldTimer value. > =========== > OpenSent: > > In this state BGP waits for an Open Message from its peer. [SNIP] > > =================== I understand. However, on this basis, we would also need a Keepalive time (since the basis on wihch this is be arrived at can be configurable). I don't know whether we do need to make such changes to the draft, and I am OK to leaving this portion as it stands. > The earlier comments on the list indicated we need to collect what > timers or Timer values were utilized in the FSM. I believe Alex > made this comment last January. (the list archives will have > this information.) > > So, I don't see any way to The DelayBgpOpenTimer is missing from the session attributes list, how about including it ? > > Comment 3: optional parameters specified > > Please suggest text. We have the following optional parameters: > > 1) Delay Open > 2) Accept unconfigured peers > 3) automatic start [Event 3] > 4) passive TCP establishment (optional) > 4 + 5 = Event 5 > 5) BGP_stop_flap (BGP Peer oscillation) [Event 6] > 6)automatic stop [Event 7] > 7)Collision detect in Established State > > If you want some of the optional parameters specified, let's > catch all. Please suggest text and location. It could > go in 8.1. My view is this can be split into two categories: Optional attributes exist per session, and optional attributes for the entire system. How about adding the following to section 8 (immediately before 8.1) % Optional session attributes that a implementation may support % for a session are: % % 1) DelayOpen flag % 2) DelayOpen Timer % 3) Perform automatic start % 4) Passive TCP establishment % 5) Perform automatic stop % 6) BGP Perform % % % Implementations may also support the following optinal attributes % that modify the behavior of the state machine % % 1) Accept un-configured peers % 2) Perform collision detect in Established state I guess it is possible to have some of these things listed under both categories. Am I opening up a can of worms here ? :-) > -------------------------------- > Comment 4 > > Old Text: > Event 1: Manual Start > Definition: Administrator manually starts peer connection. > > ... > Event 2: Manual Stop > Definition: Local system administrator manually stops the > peer connection. > New Text: > Event1: Manual start > Definition: Local system administrator manually starts the > bgp connection. > > ... > Event 2: Manual stop: > Definition: Local system administrator manually stops the > bgp connection. > > agree? Agreed. > --------------- > Comment 5: > > 5) Event3, Event5, Event6 and Event7 > > 5a) > ========= > Event 3/6/7 - cisco indicated the automatic stop, see the NANOG > archives for a discussion of the Bad Route > > http://www.merit.edu/mail.archives/nanog/2001-06/msg00805.html > > The "event" sequences are new with this version of the FSM > to identify particular events by number. I am unclear on what this issue is! > 5b > ====== > > Can we give examples of these events, as implemneted by curent > implementations ? This will be of help to new implementaions. > > An example of Auto Stop is already present in the Update event > handling section, and we can move that to this place as well. > > I believe that examples are outside the scope of this specification. > If you have text clarifying the event descriptions, please send it in. Well, the current text in the established state already gives an example for an automatic start event. I was suggesting that such examples be moved to the event description section. I also think adding text to clarify the intent behind optional events, and giving examples would help, especially for new implementors. However, if you feel this is outside the scope of the RFC, we can leave things as is. > > > ======================================================= > > 6) Event 4,5 Definition > > > Definition: Administrator manually start the peer > > Connection, but has the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing the connection. > > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > > >My understanding is this is used to indicate a peer configuration > >where we start the state machine, and wait for an incoming connection > >request from the configured peer. Can we re-phrase this section to > >indicate this. > > >How about something like: > > > >% The passive flag indicates that the state machine will wait for > >% specified peer to initiate a connection with the local system. If > >% this does not happen within a specific time (hold time), the local > >% system will then also attempt to initiate connection with the > >% specified peer. > > The text in 8.2.1.1 indicates the definition of the passive flag. Yes. > > 6a) > ========== > My understanding of your text is that you want to replace in both > sets of text: > > "The passive flag indicates the peer will listen prior to > establishing a connection". > > with: > > "The passive flag indicates that the state machine will wait for the > specified peer to initiate a connection with a local system. > > The problem with this sentence is that in the "unconfigured" case > the phrase "specified" peer is confusing. I think the original text > is clearer. I am ok to leave the text as is, based on text in 8.2.1.1. However, I do not quite understand your point. In the case of an un-configured peer, the local system cannot 'establish a connection', so is the unconfigured peer case still valid in this context ? > 6b) > ========== > If this does not happen within a specific time (hold time), the local > system will then also attempt to initiate (a) connection with the > specified peer. > > My comments: Again, the "specified peer" term is confusing. Also, > the 2nd half of the statement mixes the actions of the state machine > with the events. I believe this muddies the text instead of > clarifying it. Please refer to my comments for 6a. I think a reason for this is the requirement (in 8.2.1) that a FSM be established a-priori for "accepting each incoming TCP connection for which the peer has not beet identified". Is the intent to establish a FSM session for accepting TCP connections from un-configured peers ? If so, I am of the opinion that this is purely an implementation issue, and need not be a requirement in the spec. > > ================= > Comment 7 - > > Event 6 - (not specified, please confirm) > > > Won't a variant of this with bgp_stop_flap option set be requried ? > > We can also achieve the same by using the bgp_stop-Flap option as a > > flag that is provided as an input to the state machine. > > see comment number 3. bgp_stop_flap is an option. > > This state machine description has many events to describe the > mixture of options and actions. I believe it has a better clarity > than testing lots of flags within a state. > > A proof in case, is that many people are making comments about > specific transitions. I tink I should have phrased my question much better. We already have Event 3 - Automatic Start Event 5 - Automatic start with bgp_stop_flap option set To make things consistent, shouldn't we either a) Add 3 new events : 1) Manual start with bgp_stop flap option set 2) Manual start with passive TCP establishment and bgp_stop_flap option set 3) Automatic start with passive TCP establishment and bgp_stop_flap option set or b) Remove Event 6, and rely on a flag to tell us wether peer flap damping is to be performed for the session or not. > > ============ > Comment 8 - > > > 8) Event 5 Definition > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > > Is it correct that this event will occur only when we want > to restart > > a connection (after it had been terminated due to some reason besi de > > administrative action) that we had accepted from an > unconfigured peer ? > > The automatic start function is an implementation specific mechanism. > This text does not seek to restrict it in any fashion. Well, for new implementors, such clarifications will help undertsand how to generate and make use of this event. Again, I am OK with leaving text as is. > > ================ > 9) Timer Events Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > > > Event9: connect retry timer expires > > > > Definition: An event triggered by the expiration of=20 > > Event10: hold time expires > > > > Definition: An event generated when the HoldTimer=20 > > We can use any one of these terms (trigerred by/generated > when) for all > timer event definitions to make it look more consistent. > > agree: Will use "generate" > > Event 8: An event generated when the Idle Hold timer expires. > Event 9: An event generated when the ConnectRetry timer expires. > Event 10: An event generated when the Hold timer expires. > Event 11: An event generated when the Keepalive timer expires > Event 12: An event generated when the Delay BGP Open timer expires. > > Please confirm that this change to the definitions is agreeable to > you. Agreed. I think we have consensus on this issue, unless anyone else has any comments on this. > > ================ > 10) Event 8 Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > Hold Timer is only used when persistent =20 > > BGP oscillation damping functions are=20 > > enabled. =20 > > > > Status: Optional. Used when persistent > > BGP peer oscillation damping functions > > are enabled.=20 > > > The status repeats the senetnce in the definition. Can we remove the > sentence from the Status, or replace with a different definition. > > Do you think the defenition below would fit ? > > % An event trigerred by the expiry of the Idle Hold timer, indicating > % that the session has completed waiting for a back-off period to > % prevent flap, and may now proceed to take any further systemm > % initiated action for this session. > > text change suggestion for 1st sentence that I'll agree to. > > An event trigerred by the expiry of the Idle Hold timer, indicating > that the session has completed waiting for a back-off period to > prevent bgp peer oscillation. > > Please confirm that you will agree to this. Agreed. I think we have consensus on this issue, unless anyone else has any comments on this. Siva (siva@ctd.hcltech.com)> Original text of comments 1-10 follow: > > ------- Message 4 > > Date: Fri, 29 Nov 2002 19:46:06 +0530 > From: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com> > To: skh@nexthop.com, Yakov Rekhter <yakov@juniper.net> > cc: skh@ndzh.com > Subject: Comments on draft 18 > > This message is in MIME format. Since your mail reader does > not understand > this format, some or all of this message may not be legible. > > - ------_=_NextPart_000_01C297B1.DB0F1700 > Content-Type: text/plain; > charset="iso-8859-1" > > Hello, > > Attached are some comments that I have on draft-ietf-idr-bgp4-18. > > Please have a look at these and tell me if the same can > be raised to the > mailing list as well. > > Thanks, > Siva > (siva@ctd.hcltech.com) > > > > - ------_=_NextPart_000_01C297B1.DB0F1700 > Content-Type: text/plain; > name="BGP_Issues.txt" > Content-Transfer-Encoding: quoted-printable > Content-Disposition: attachment; > filename="BGP_Issues.txt" > > All quoted sections of the original document have been > prefixed by a '> = > ' > All lines of proposed text have been prefixed by a '% ' > > > 1) Performing of peer oscillation damping: > > Since this feature is going to be added in a new draft, and its > addition will change the operation of the state machine, can we remove > all mention of it in the state machine ? As part of this removal, can > we also remove the IdleHold Timer from the FSM since it is not useful > in the absence of peer oscillation damping ? > > The draft that describes this procedure can then describe the change > in the state machine required to do this. > > > 2) Session Attributes: > > > Session Attributes required for each connection are;=20 > > > 1) state > > 2) Connect Retry timer > > 3) Hold timer > > 4) Hold time > > 5) Keepalive timer > > Why have a Hold Timer and a Hold Time ? Can we replace this with just > Hold timer ? > > Can we also add the following session attributes: > > a) DelayBgpOpenTimer > b) IdleHold Timer (in case we choose not to remove this from the bas = > FSM) > > Can we also add the following flag to the session attributes: > a) DelayOpen Flag > > 3) All sessions attributes: > > There are some attributes that are common to all sessions of the = > system. > Can we call these out as well: > > * Accept Connections from unconfigured peers (Enabled/Disabled) > * Peer Oscillation Dampening (Enabled/Disabled) (In case we choose > not to remove it from base spec) > > 4) Description for Event1/Event2 > > > Event1: Manual start=20 > > > > Definition: Administrator manually starts peer > > connection > > > Event2: Manual stop > > > Definition: Local system administrator manually=20 > > stops the peer connection. > > We can make these look consistent by using the Term Administrator or > Local System administrator in both cases. > > 5) Event3, Event5, Event6 and Event7 > > Can we give examples of these events, as implemneted by curent > implementations ? This will be of help to new implementaions. > > An example of Auto Stop is already present in the Update event > handling section, and we can move that to this place as well. > > 6) Event 4,5 Definition > > > Definition: Administrator manually start the peer > > Connection, but has the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing the connection. > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > My understanding is this is used to indicate a peer configuration > where we start the state machine, and wait for an incoming connection > request from the configured peer. Can we re-phrase this section to > indicate this. > > How about something like: > > % The passive flag indicates that the state machine will wait for > % specified peer to initiate a connection with the local system. If > % this does not happen within a specific time (hold time), the local > % system will then also attempt to initiate connection with the > % specified peer. > > 7) Event 4,5 > > Won't a variant of this with bgp_stop_flap option set be requried ? > We can also achieve the same by using the bgp_stop-Flap option as a > flag that is provided as an input to the state machine. > > 8) Event 5 Definition > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > Is it correct that this event will occur only when we want to restart > a connection (after it had been terminated due to some reason beside > administrative action) that we had accepted from an > unconfigured peer ? > > 9) Timer Events Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > > > Event9: connect retry timer expires > > > > Definition: An event triggered by the expiration of=20 > > Event10: hold time expires > > > > Definition: An event generated when the HoldTimer=20 > > We can use any one of these terms (trigerred by/generated > when) for all > timer event definitions to make it look more consistent. > > 10) Event 8 Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > Hold Timer is only used when persistent =20 > > BGP oscillation damping functions are=20 > > enabled. =20 > > > > Status: Optional. Used when persistent > > BGP peer oscillation damping functions > > are enabled.=20 > > The status repeats the senetnce in the definition. Can we remove the > sentence from the Status, or replace with a different definition. > > Do you think the defenition below would fit ? > > % An event trigerred by the expiry of the Idle Hold timer, indicating > % that the session has completed waiting for a back-off period to > % prevent flap, and may now proceed to take any further systemm > % initiated action for this session. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id QAA12713 for <idr-archive@nic.merit.edu>; Fri, 13 Dec 2002 16:51:55 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id E002A912FA; Fri, 13 Dec 2002 16:51:30 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id AD146912FC; Fri, 13 Dec 2002 16:51:30 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 43F3F912FA for <idr@trapdoor.merit.edu>; Fri, 13 Dec 2002 16:51:29 -0500 (EST) Received: by segue.merit.edu (Postfix) id 2A0705E00B; Fri, 13 Dec 2002 16:51:29 -0500 (EST) Delivered-To: idr@merit.edu Received: from nemesis.systems.pipex.net (nemesis.systems.pipex.net [62.241.160.8]) by segue.merit.edu (Postfix) with ESMTP id D31425E005 for <idr@merit.edu>; Fri, 13 Dec 2002 16:51:28 -0500 (EST) Received: from tom3 (userbf91.uk.uudial.com [62.188.142.112]) by nemesis.systems.pipex.net (Postfix) with SMTP id A908816008BB3; Fri, 13 Dec 2002 21:51:21 +0000 (GMT) Message-ID: <004b01c2a2f1$bb97c8e0$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: "Susan Hares" <shares@nexthop.com>, "idr" <idr@merit.edu> Cc: "Jeffrey Haas" <jhaas@nexthop.com> Subject: Event 19 in Open Sent state was Re: bgp18 WG Last Call fsm missing keepalive Date: Fri, 13 Dec 2002 21:46:45 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Sue You asked if event 19 (Open with OpenDelay timer running) was possible in OpenSent state; I gave a lengthy reply (below) to the effect yes it could because the OpenDelay timer did not always get stopped but the timer should be stopped in which case the event would not happen. Reading your responses to Siva , I see you include stopping the Open Delay timer in the action 'release all BGP resources' when going to Idle (which I missed seeing earlier in the year). That eliminates most but not all of the possibilities I mentioned. I now believe we would need to add the action 'stop OpenDelay timer' for Connect state event 17 (currently defined as going to Active) event 9 (stays in Connect state) in order to stop event 19 in Open Sent Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Tom Petch <nwnetworks@dial.pipex.com> To: Susan Hares <shares@nexthop.com>; idr <idr@merit.edu> Cc: Jeffrey Haas <jhaas@nexthop.com> Date: 05 December 2002 19:02 Subject: Re: bgp18 WG Last Call fsm missing keepalive >Sue <snip> >2) complicated. As the fsm text stands, Event 19, along with a host >of others, takes us back from Open Sent to Idle (I assume on the >grounds this is an error condition) which seems very reasonable. > >But ...in quite a few places, such as Connect state events 2, >7,8,9,10,11, 17, 18, 20 thru 27, we do not stop the OpenDelay timer >when going to Idle or Active so we could then go from eg Idle with >Manual start [event 1] to Connect to Open Sent all before the >OpenDelay timer expires in which case event 19 can occur validly in >Open Sent - obscure but possible. (This is also true with Active state >and events 2, 17 and the default list at the end). > >But I think this is an error, and that when exiting Connect state or >Active state as listed above, we should have an additional action to >stop the OpenDelay timer in which case event 19 in Open Sent becomes >an error condition (again). > >But but but as ever, I cannot speak with authority for implementations >and so if implementations do not stop the OpenDelay timer when exiting >as above, then Event 19 is valid in Open Sent state - obscure but >possible (again). > >My wish is to add the extra action, stop OpenDelay timer, for the >events listed above in Active and Connect states in the expectation >that that is what people have or should have implemented. > >Tom Petch >nwnetworks@dial.pipex.com > >-----Original Message----- >From: Susan Hares <shares@nexthop.com> >To: Tom Petch <nwnetworks@dial.pipex.com>; idr <idr@merit.edu> >Cc: Jeffrey Haas <jhaas@nexthop.com> >Date: 04 December 2002 19:50 >Subject: RE: bgp18 WG Last Call fsm missing keepalive > > >Tom: > <snip> >2) Do you think that Event 19 is possible in the Open Sent state? > >Please answer this separately. > >Thanks for all the wonderful review of the State machine. >I'm working my way through the comments. > >Sue > >- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id OAA04346 for <idr-archive@nic.merit.edu>; Fri, 13 Dec 2002 14:31:54 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 7449C912F7; Fri, 13 Dec 2002 14:31:28 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id CF8EF912F8; Fri, 13 Dec 2002 14:31:27 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 7C083912F7 for <idr@trapdoor.merit.edu>; Fri, 13 Dec 2002 14:31:26 -0500 (EST) Received: by segue.merit.edu (Postfix) id 33C155DFCC; Fri, 13 Dec 2002 14:31:01 -0500 (EST) Delivered-To: idr@merit.edu Received: from workhorse.fictitious.org (workhorse.fictitious.org [209.150.1.230]) by segue.merit.edu (Postfix) with ESMTP id 609575DF45 for <idr@merit.edu>; Fri, 13 Dec 2002 14:31:00 -0500 (EST) Received: from workhorse.fictitious.org (localhost.fictitious.org [127.0.0.1]) by workhorse.fictitious.org (8.9.3/8.9.3) with ESMTP id OAA52205; Fri, 13 Dec 2002 14:30:50 -0500 (EST) (envelope-from curtis@workhorse.fictitious.org) Message-Id: <200212131930.OAA52205@workhorse.fictitious.org> To: "Natale, Jonathan" <JNatale@celoxnetworks.com> Cc: isp-routing@isp-routing.com, idr@merit.edu Reply-To: curtis@fictitious.org Subject: Re: deterministic-med In-reply-to: Your message of "Fri, 13 Dec 2002 12:37:00 EST." <1117F7D44159934FB116E36F4ABF221B02C7C715@celox-ma1-ems1.celoxnetworks.com> Date: Fri, 13 Dec 2002 14:30:50 -0500 From: Curtis Villamizar <curtis@fictitious.org> Sender: owner-idr@merit.edu Precedence: bulk In message <1117F7D44159934FB116E36F4ABF221B02C7C715@celox-ma1-ems1.celoxnetwor ks.com>, "Natale, Jonathan" writes: > Is any reason why someone would want 'non-deterministic-med' behavior, I do > not. Maybe I missed something. > > thnx I don't agree with it but I've heard the argument that if there are two or more routes and one route flaps they'd rather just lock in one that doesn't flap. I haven't heard that argument in a while. The obvious counter argument is persistent route loops make customers very unhappy and route damping does the job. Curtis Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id MAA27934 for <idr-archive@nic.merit.edu>; Fri, 13 Dec 2002 12:37:27 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 9811691209; Fri, 13 Dec 2002 12:37:03 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 61D72912EE; Fri, 13 Dec 2002 12:37:03 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 3931891209 for <idr@trapdoor.merit.edu>; Fri, 13 Dec 2002 12:37:02 -0500 (EST) Received: by segue.merit.edu (Postfix) id 14AB95DDE0; Fri, 13 Dec 2002 12:37:02 -0500 (EST) Delivered-To: idr@merit.edu Received: from celox-ma1-ems1.celoxnetworks.com (celox-ma1-imap1.celoxnetworks.com [12.40.60.233]) by segue.merit.edu (Postfix) with ESMTP id D0F255DDB2 for <idr@merit.edu>; Fri, 13 Dec 2002 12:37:01 -0500 (EST) Received: by celox-ma1-ems1.celoxnetworks.com with Internet Mail Service (5.5.2653.19) id <YN1W9ASW>; Fri, 13 Dec 2002 12:37:01 -0500 Message-ID: <1117F7D44159934FB116E36F4ABF221B02C7C715@celox-ma1-ems1.celoxnetworks.com> From: "Natale, Jonathan" <JNatale@celoxnetworks.com> To: isp-routing@isp-routing.com, idr@merit.edu Subject: deterministic-med Date: Fri, 13 Dec 2002 12:37:00 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain Sender: owner-idr@merit.edu Precedence: bulk Is any reason why someone would want 'non-deterministic-med' behavior, I do not. Maybe I missed something. thnx Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id SAA01598 for <idr-archive@nic.merit.edu>; Thu, 12 Dec 2002 18:23:01 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 6C07D912E7; Thu, 12 Dec 2002 18:22:34 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 3B6B1912E8; Thu, 12 Dec 2002 18:22:34 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 0307E912E7 for <idr@trapdoor.merit.edu>; Thu, 12 Dec 2002 18:22:32 -0500 (EST) Received: by segue.merit.edu (Postfix) id D95675DE1A; Thu, 12 Dec 2002 18:22:32 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 7E6D25DDAB for <idr@merit.edu>; Thu, 12 Dec 2002 18:22:32 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gBCNMVV55750; Thu, 12 Dec 2002 18:22:31 -0500 (EST) (envelope-from jhaas@jhaas.nexthop.com) Received: from jhaas.nexthop.com (jhaas.nexthop.com [64.211.218.31]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gBCNMSC55743; Thu, 12 Dec 2002 18:22:28 -0500 (EST) (envelope-from jhaas@jhaas.nexthop.com) Received: (from jhaas@localhost) by jhaas.nexthop.com (8.11.3nb1/8.11.3) id gBCNMS709777; Thu, 12 Dec 2002 18:22:28 -0500 (EST) Date: Thu, 12 Dec 2002 18:22:28 -0500 From: Jeffrey Haas <jhaas@nexthop.com> To: Susan Hares <shares@nexthop.com> Cc: idr@merit.edu Subject: Re: Comments 30-36 from Siva Message-ID: <20021212182228.R6330@nexthop.com> References: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AB0E@aa-exchange1.corp.nexthop.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AB0E@aa-exchange1.corp.nexthop.com>; from shares@nexthop.com on Thu, Dec 05, 2002 at 04:18:18PM -0500 X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Sue, On Thu, Dec 05, 2002 at 04:18:18PM -0500, Susan Hares wrote: > 33) Handling Event 18 in the OpenSent state > > >Why do we start the Keepalive timer at this stage ? Isn't it sufficient = > >to do so when we move into Established state ? > > An earlier comment from Tom (and you) requested the following: > > <---Open > [Open sent state] > > Open--> > [Event 18] > > <---Open > <---Keepalive > [Action from Event 18 in Open Sent] > [Open Confirm] > Keepalive -> [Event 25] > > [established] I'm assuming the second OPEN going from right to left is a typo. If it isn't, thats a FSM error to the peer on the left. Theoretically, an implementation that utilizes its keepalive timer to send the first keepalive to transition to Established is still interoperable. However: o Keepalives can be disabled by negotiating hold time of zero o We really shouldn't need to restart the Keepalive timer. If there is a delay in the keepalive that transitions from OpenConfirm to Established, its due to the transport connection. It should be reliable and it *should* get through. If it doesn't, there's other problems and the hold timer for the peer on the right should do the Right Thing and drop the connection. > What do implementations do? We'll have to query implementations. GateD at least waits to enter the Established state prior to starting the KeepAlive timer. -- Jeff Haas NextHop Technologies Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id SAA00799 for <idr-archive@nic.merit.edu>; Thu, 12 Dec 2002 18:08:22 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id BBDFA912E4; Thu, 12 Dec 2002 18:07:59 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 8D4B0912E6; Thu, 12 Dec 2002 18:07:59 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 366D0912E4 for <idr@trapdoor.merit.edu>; Thu, 12 Dec 2002 18:07:57 -0500 (EST) Received: by segue.merit.edu (Postfix) id CAACF5DDFA; Thu, 12 Dec 2002 18:07:57 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 6AA055DDD2 for <idr@merit.edu>; Thu, 12 Dec 2002 18:07:57 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gBCN7uo55062; Thu, 12 Dec 2002 18:07:56 -0500 (EST) (envelope-from jhaas@jhaas.nexthop.com) Received: from jhaas.nexthop.com (jhaas.nexthop.com [64.211.218.31]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gBCN7mC55049; Thu, 12 Dec 2002 18:07:48 -0500 (EST) (envelope-from jhaas@jhaas.nexthop.com) Received: (from jhaas@localhost) by jhaas.nexthop.com (8.11.3nb1/8.11.3) id gBCN7mt09636; Thu, 12 Dec 2002 18:07:48 -0500 (EST) Date: Thu, 12 Dec 2002 18:07:48 -0500 From: Jeffrey Haas <jhaas@nexthop.com> To: Susan Hares <shares@nexthop.com> Cc: idr@merit.edu Subject: Re: Comments 21-30 Message-ID: <20021212180748.Q6330@nexthop.com> References: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AB0C@aa-exchange1.corp.nexthop.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AB0C@aa-exchange1.corp.nexthop.com>; from shares@nexthop.com on Thu, Dec 05, 2002 at 04:21:00PM -0500 X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk On Thu, Dec 05, 2002 at 04:21:00PM -0500, Susan Hares wrote: > 23) Handling events 20, 21 in the Connect State and Active State > > We need to consider the case where we receive events 20 (message > header error) and 21 (Open message error) when the delay timer is > running. > > Since the connection has been established at this point, we need to > send a Notification message and then terminate the connection. > > Alternative comments: > > 1) We have not sent an Open statement. > 2) Why do we have to send an Notification? I see no justification > for it. > > Suggestion: > Do you have implementations that send notification? Do > you know of others that don't. This points out a potential issue. >From section 4.2 of -18: : 4.2 OPEN Message Format : : After a TCP is established, the first message sent by each side is an : OPEN message. If the OPEN message is acceptable, a KEEPALIVE message : confirming the OPEN is sent back. Once the OPEN is confirmed, UPDATE, : KEEPALIVE, and NOTIFICATION messages may be exchanged. This text implies that NOTIFICATIONs can only be sent once we have sent an open and then a keepalive, generally meaning we're in the Established state. Anyone suggestions for modifying the wording? Section 6.1 (Message header error) is one situation that implies that a NOTIFICATION can be sent without sending even an OPEN message. Note that since the base FSM implies that we send an OPEN message immediately when we have a completed trasnport connection, we SHOULD be in at least OpenSent. However, the DelayOpen timer means that we MAY send a NOTIFICATION when we are in the Connect state. GateD, at least, will not send a NOTIFICATION without first sending an OPEN. We need to pick one: You can send NOTIFICATIONS before OPEN or before OPEN if the OpenDelay timer is running. However, we MUST fix the text above. > ============================= > > 26) Handling Event 17 in The connect state > > > If the transport connection fails (timeout or transport > > disconnect) [Event17], the local system: > > - changes its state to Active.=20 > > >If the transport connection fails when the BGP Open Delay timer is > > running, should we still be going into the Active state ? > > see mail thread on eventse 13-17 In this particular case, I think the issue is separate from the issues for events 13-17 since this isn't particular to how deep the BGP implementation meddles in the TCP implementation. If we are in the Connect state, because we have an incoming transport connection that has completed, but we have the OpenDelay timer running and the transport connection is closed, we can simply drop into Active after resetting the ConnectRetry timer and clearing the OpenDelay timer (if set/exists). In the case of an unconfigured peer, we can discard the FSM instance. > ========================================= > > 29) Handling Event 2 in Active state > > >> A manual stop event[Event2], the local system: > >> - Sends a notification with a Cease, > >> - drops the Transport connection,=20 > > > >These two actions are possible only if a transport connection had = > > already been established. How about changing the text to > > > > % - If a transport connection had been successfully established > > % - Send a Notification with a Cease > > % - Drop the Transport Connection > > How about this text: > > A manual stop event [Event 2], the local system > - Drop the TCP connection, > - Release all BGP resources, > - resets the connection retry timer [sets to zero], > - goes to Idle. I'm rather confused. Under exactly what circumstances can we be in the Active state and have an active TCP connection at all? Ditto for having any BGP resources? Going to Idle is fine. -- Jeff Haas NextHop Technologies Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id RAA28177 for <idr-archive@nic.merit.edu>; Thu, 12 Dec 2002 17:19:45 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 83A5E912E5; Thu, 12 Dec 2002 17:16:30 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id C5BA3912E4; Thu, 12 Dec 2002 17:16:28 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 27747912E2 for <idr@trapdoor.merit.edu>; Thu, 12 Dec 2002 17:16:24 -0500 (EST) Received: by segue.merit.edu (Postfix) id 27C0D5DEC5; Thu, 12 Dec 2002 17:16:01 -0500 (EST) Delivered-To: idr@merit.edu Received: from shockwave.systems.pipex.net (shockwave.systems.pipex.net [62.241.160.9]) by segue.merit.edu (Postfix) with ESMTP id 7E60D5DE3C for <idr@merit.edu>; Thu, 12 Dec 2002 17:16:00 -0500 (EST) Received: from tom3 (userbh03.uk.uudial.com [62.188.142.222]) by shockwave.systems.pipex.net (Postfix) with SMTP id 69CAB16001ADD; Thu, 12 Dec 2002 22:15:54 +0000 (GMT) Message-ID: <001c01c2a22b$feb0faa0$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com>, "Susan Hares" <shares@nexthop.com>, "Yakov Rekhter" <yakov@juniper.net>, <idr@merit.edu> Cc: "Venu Kumar G - CTD, Chennai." <venug@ctd.hcltech.com> Subject: Re: Comments 11-20 Date: Thu, 12 Dec 2002 22:13:53 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MIMEOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Comments ** in line Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Sivananda Ramnath - CTD, Chennai. <siva@ctd.hcltech.com> To: Susan Hares <shares@nexthop.com>; Yakov Rekhter <yakov@juniper.net>;idr@merit.edu <idr@merit.edu> Cc: Venu Kumar G - CTD, Chennai. <venug@ctd.hcltech.com> Date: 10 December 2002 18:09 Subject: RE: Comments 11-20 > Revised summary: > > 11 - No consensus, but current text can stand as is > any comments from others ? ** these are two distinct timers, two distinct events (KeepAlive 11 and Hold 10) so keep the existing text > > 12 - text changed, potential consensus ** Open Delay timer in any new text please > 13a - text changed, potential consensus > 13b - no change, potential consensus **I found and find ** Transport Connection Indication & valid remote peer **verbose and clumsy to read and write and suggested/suggest ** 13 Transport connection valid indication ** 14 Transport connection invalid indication **at most, even cutting the names down to three words if possible. ** There remains plenty of scope for verbosity in the textual description although I think we should beware of nailing it down too tightly. eg Invalid in a security conscious system might come to mean this address has been added to a black list on account of its recent malicious behaviour, perhaps dynamically blocking some source port numbers on account of denial of service attacks. While valid in a lax implementation could mean anything goes, even port 179. I am ok with the existing description. > > > 14 - additional comments, no consensus > Note: No consensus on one part only > 15 - Consensus, author's choice. > 16 - Agree to your comments. > Have clarified original comment, > proposed some text. ** Open Delay timer in any new text please > 17 - consensus ** Open Delay timer in any new text please > 18 - text revised, potential consensus > 19 - no change, potential consensus > 20 - text changed, potential consensus > > >> Here's the summary of 11-20 >> >> comment: >> >> 11 - no consensus, no text supplied >> >> note: usage varies from my understanding >> of common usage. >> >> 12 - consensus, no text supplied >> see comment 9 and proposed text below. >> >> 13a - no consensus, see alternate text >> 13b - no consensus, implementation specific detail >> >> 14 - no consensus, disagree with comment >> policy issues, outside scope of document >> >> 15 - no consensus, see alternative text. >> 16 - no consensus, see my proposed text. >> >> You did not propose text >> See alternate text for events 13-17 >> Also note, that I consider that >> events 13 and 14 should go to >> optional. >> >> 17 - consensus, agree to change in text. >> 18 - agree needs to be fixed, see my alternative text. >> 19 - further clarity needed, >> please respond to question >> >> 20 - agree, text will be changed >> >> >> Sue >> >> ===================== >> >> 11) Event 10 (Hold timer) >> >> We use the hold timer for two purposes >> >> * Waiting for an open message (with a default value of 240 seconds) >> * Waiting for Keepalives (with a default value of 90 seconds) >> >> Can we use two different timers (or at least call them two different >> timer events) ? >> >> Siva: I do not believe that most implementatios >> utilize two timers. Can you indicate implementations >> that do? > > Implementations can certainly use a single timer, since at any time > at most only one of these two timers will need to be running. > > The reason for the comment is that conceptually these are two > different timers: > > One is a standard (or configurable) value, and is used to wait for an > Open Message. > > The other is negotiated, and is used to wait for Keepalives. > > I feel that representing these as two different events will make the > state machine look better. Of course, nothing prevents an implementation > from achieving this by using the same timer, or any other means as long > the same effect is realized; so existing implementations will not be > affected in any fashion. > > I am not very particular on this issue, since things will work fine > with the current means as well, do others have any comments on > modeling this as two separate events in the state machine ? > >> >> ============================================================== >> ========= >> 12) Event 12 >> >> > Event12: DelayBGP Open timer expires [changed] >> > >> > Definition: A timer that delays sending of the BGP >> > Open message for n seconds after the >> > Transport connection has been completed.= >> >> >The Definition seems to better define what the timer is, >> rather than = >> > what >> > its expiry means. Can we change it to reflect this ? >> >> new text: see comment on 9 on Timer definition, >> and earlier change. >> >> Event 12: An event generated when the Delay BGP Open timer expires. >> ========================================= >> > > Ok. > Can we add any text on the need or use of a Delay Open ? > > >> 13) Event 13, 14 definition >> >> > Event13: Transport Connection Indication & valid remote >> peer[changed] >> > >> > Definition: Event indicating that TCP connection >> > request with a valid source IP address and TCP >> > port, and valid destination IP address >> > and TCP Port. The definition of >> > invalid Source, and invalid Destination= >> > IP address is left to the implementation. >> >> >> a) How about changing this slighty to: >> % Definition: Event indicating receiving a TCP connection >> % request with a valid source IP address and TCP >> % port, and valid destination IP address= >> % and TCP Port. The definition of invalid >> My completion of your sentence: >> source, and invalid destination >> IP address >> is left to the implementation. >> >> How about: >> Definition: Event indicating the reception of a TCP connection >> request with a valid source IP address and TCP >> port, and valid destination IP address and >> TCP Port. The definition of invalid >> source address >> and port and invalid destination address >> is left to the implementation. >> >> >> Please comment if you agree with this change. > > This looks good. > >> >> b) Is it correct that Destination IP address validation is >> used when we >> do not wish to accept BGP connections on one of the local IP addresses >> (either because we wish to accept such addresses on only our loopback >> interfaces, or because the address to which we received the connection >> was not the optimal interface with which to speak to the sender) ? If >> so, can we add this to the description ? >> >> >> No, this is an implementation dependent change. > > Ok. Again, if there are any validations that are considered > desirable, adding this to the draft text as a suggestion to > implementers would be useful. > > We can leave things as is if you feel that such details should > be left out of the draft. > > >> ============================================================== >> =================== > > >> >> 14) Event 13,14 >> >> > Event13: Transport Connection Indication & valid remote >> peer[changed] >> > >> > Definition: Event indicating that transport >> > Request valid source IP address and TCP=20 >> > port, and valid destination IP address=20 >> > and TCP Port. The definition of=20 >> > invalid Source, and invalid Destination=20 >> > IP address is left to the implementation. >> > >> > a) I am unclear as to what validation of source and >> destination port = >> > mean. >> > >> > The source port's value should be of no concern to connection = >> > receipient system. >> >> The definition of "valid" is an implementation policy issue. >> Validating is the process of checking to make sure it is valid >> for the implementation's policy. > > Ok. > >> >> > The destination port cannot be invalid, since BGP would not >> receive the >> > connection request in such a situation. >> >> Again, this is a policy issue. We are providing hooks for >> policy to be engaged. Policy is outside the scope of this >> document. > > Isn't this more than just a policy issue ? It seems difficult > to identify a connection destined to a port other than to > the one on which BGP is running as an error event for BGP. > So, I don't understand how a policy can be applied in this > regard. > >> >> > b) I think it will be useful to state what we consider to >> be an invalid >> > IP address. >> >> Again, it is policy issue which is implementation dependent. > > Ok. > >> >> ==================== >> 15) Event 17 >> >> > Event17: TCP connection fails >> >> > This event can occur even when the transport connection is closed by >> > the other end. Since this does not reflect a 'failure ', >> can we change >> > the event name to >> > >> > % Event17: TCP connection termination >> >> Discussion: It both terminates from the remote site >> and can "timeout" - fail. >> >> Suggestions? I can use "disconnect", what do you think. > > That looks fine too. I don't know what the best words > to describe this are. Also, the description for this > event does a good job of explaining this, so either > the original text (connection fails) or your suggestion > (disconnect) can be used. > This really should not have been raised as a comment. :-) > >> ========================== >> >> 16) Events related to BGP Messages >> >> Can we make the definition style across these events consistent ? >> >> thanks for the comment. >> proposed text below, please agree on it. > > Yes, the text below looks fine. > > I should have phrased the original comment much better, I was > referring to consistency across the following: Events 18, 19, > 20, 21 vs. Events 23, 24, 25, 26, 27. > > This is a very minor issue, so not making these changes is fine > too. The way the latter events are described > are different from the former. I suggest adopting the style for > Events 23-27 for Events 18-21. > > As per this, the following changes would have to be made: > >> Event18: BGPOpen >> >> Definition: An event indicating that a valid Open >> message has been received. > > with > >% Event18: BGPOpen >% >% Definition: An event is generated when a valid Open >% message has been received. > >> Event19: BGPOpen with BGP Delay Open Timer running >> >> Definition: An event indicating that a valid Open >> message has been successful >> established for a peer that is >> currently delaying the sending of an >> BGP Open message. > > with > >% Event19: BGPOpen with BGP Delay Open Timer running >% >% Definition: An event is generated when a valid Open >% message has been received for a peer that >% is currently delaying the sending of a >% BGP Open message. > >> Event20: BGPHeaderErr >> >> Definition: BGP message header is not valid. > > with > >% Event20: BGPHeaderErr >% >% Definition: An event is generated when a received BGP >% message header is not valid. > >> Event21: BGPOpenMsgErr >> >> Definition: An BGP Open message has been received >> with errors. > > with > >% Event21: BGPOpenMsgErr >% >% Definition: An event is generated when BGP Open message >% with errors has been received. > > >> >> proposed text replacement >> ------------------------ >> Event13: TCP connection indication and valid remote peer >> >> Definition: Event indicating the local system reception of >> a TCP connection request with a valid source >> IP address and TCP port, and valid destination >> IP address and TCP Port. The definition of >> invalid source, and invalid destination >> IP address is left to the implementation. >> >> BGP's destination port SHOULD be port >> 179 as defined by IANA. >> >> TCP connection request is denoted by >> the local system receiving a TCP SYN. >> >> Status: Mandatory (Optional) >> >> >> Event14: RCV TCP connection indication with invalid source or >> destination >> >> Definition: Event indicating the local system reception of >> a TCP connection request with either >> an invalid source address or port >> number or an invalid destination >> address or port number. >> >> BGP destination port number SHOULD be 179 >> as defined by IANA. >> >> Again, a TCP connection request >> denoted by local system receiving a TCP >> SYN. >> >> Status: Mandatory (Optional) >> >> Event15: TCP connection request sent received an ACK. >> >> Definition: Event indicating the Local system's request >> to establish a TCP connection to the remote >> peer. >> >> The local system's TCP session sent a TCP >> SYN, and received a TCP SYN, ACK pair of >> messages, and Sent a TCP ACK. >> >> Status: Mandatory >> >> Event16: TCP connection confirmed >> >> Definition: Event indicates that the local system receiving >> a confirmation that the TCP connection has >> been established by the remote site. >> >> The remote peer's TCP engine sent a TCP SYN. >> The local peer's TCP engine sent a SYN, ACK >> pair, and now has received a final ACK. >> >> Status: Mandatory >> >> Event17: TCP connection fails >> >> Definition: Event indicates that the local system has >> received a TCP connection failure notice. >> >> The remote BGP peer's TCP machine could have >> sent a FIN. The local peer would respond >> with a FIN-ACK. Another alternative is that >> the local peer indicated a timeout in the >> TCP session and downed the connection. >> >> Status: Mandatory >> >> --------------------- >> >> >> 17) Event 19 Definition >> >> > Event19: BGPOpen with BGP Delay Open Timer running [changed] >> > >> > Definition: An event indicating that a valid Open=20 >> > Message has been successful=20 >> > established for a peer that is=20 >> > currently delaying the sending of an=20 >> > BGP Open message.=20 >> >> >> The definition appears a little fragmented. How about: >> >> % Definition: An event indicating that a valid OPEN >> % Message has been received for a peer >> % that has a successfully established >> % transport connection and is currently >> % delaying the seending of a BGP open >> % message >> >> >> agreed, changed in text. I will send out definition >> text for review in a separate message. > > Ok. > >> ============================================================== >> 18) Event 22 Definition >> >> a)=20 >> >> > Event22: Open collision discard >> > >> > Definition: An event generated administratively=20 >> > when a connection Collision has been=20 >> > detected while processing an incoming =20 >> > Open message. This connection has been=20 >> >> Isn't this event 'automatically' generated, since it is a system >> generated event ?=20 >> >> response: How this generated is implementation specific. The >> "administratively" is to cover policy. >> >> b) >> >> > Event 22 is an administrative could=20 >> > occur if FSM is implemented as two=20 >> > >> >> The word event is missing. How about >> >> % Event 22 is an automatic event that >> % could occur if FSM is implemented as two=20 >> >> >> I've re-written the whole definition. What do you think about: >> >> Event22: Open collision dump >> >> Definition: An event generated administratively >> when a connection collision has been >> detected while processing an incoming >> OPEN message and this connection has been >> selected to disconnected. See Section >> 6.8 for more information on collision >> detection. >> >> Event22 is an administrative based only >> implementation specific policy. This >> Event may occur if the FSM is implemented >> as two linked state machines. >> >> Please confirm that you like this. > > Agreed. > >> >> ==================================== >> 19) FSM Description: ConnectRetry Count >> >> > The Connect Retry count is updated by the FSM but never used. In the >> > absence of peer oscillation damping, will this be used to stop >> > connection establishment attempts after a certain maximum number ? >> >> Yes, this is either implementation specific or >> is it based on the peer oscillation damping draft. >> >> > Can we include the use of this counter in some place ? >> >> Connect retry counter >> 1) Will be utilized by the peer oscillation >> damping drat. >> 2) Will be included in bgp-4-mibv2-xx. >> I just check and I didn't find it. >> >> Do you still want text in the main > > We can leave the main text as is. > >> >> ==================================== >> >> 20) Handling Event 7 (Auto Stop) to Idle State processing: >> >> The handling of Event 7 is missing from the Idle State processing. Can >> we add this ? How about replacing >> >> > An manual stop event (Event2) is ignored in the Idle state. >> >> with >> >> % Manual stop (Event 2) and Auto stop (Event 7) events are ignored >> % in the Idle state >> >> we'll add that in the text. > > Ok. > >> >> ============================== >> > >Siva >(siva@ctd.hcltech.com) Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id RAA27984 for <idr-archive@nic.merit.edu>; Thu, 12 Dec 2002 17:16:54 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 7B7FD912E3; Thu, 12 Dec 2002 17:16:30 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 85DC3912E2; Thu, 12 Dec 2002 17:16:28 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 9310B912E3 for <idr@trapdoor.merit.edu>; Thu, 12 Dec 2002 17:16:24 -0500 (EST) Received: by segue.merit.edu (Postfix) id DD5A75DE12; Thu, 12 Dec 2002 17:16:04 -0500 (EST) Delivered-To: idr@merit.edu Received: from shockwave.systems.pipex.net (shockwave.systems.pipex.net [62.241.160.9]) by segue.merit.edu (Postfix) with ESMTP id 21DB05DE3C for <idr@merit.edu>; Thu, 12 Dec 2002 17:16:04 -0500 (EST) Received: from tom3 (userbh03.uk.uudial.com [62.188.142.222]) by shockwave.systems.pipex.net (Postfix) with SMTP id 01A0C160009C6; Thu, 12 Dec 2002 22:16:00 +0000 (GMT) Message-ID: <001d01c2a22c$010d3a20$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com>, "Susan Hares" <shares@nexthop.com>, "Yakov Rekhter" <yakov@juniper.net>, "idr" <idr@merit.edu> Cc: "Venu Kumar G - CTD, Chennai." <venug@ctd.hcltech.com> Subject: Re: Comments 11-20 Date: Thu, 12 Dec 2002 22:14:39 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MIMEOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Comments ** in line Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Sivananda Ramnath - CTD, Chennai. <siva@ctd.hcltech.com> To: Susan Hares <shares@nexthop.com>; Yakov Rekhter <yakov@juniper.net>;idr@merit.edu <idr@merit.edu> Cc: Venu Kumar G - CTD, Chennai. <venug@ctd.hcltech.com> Date: 10 December 2002 18:09 Subject: RE: Comments 11-20 > Revised summary: > > 11 - No consensus, but current text can stand as is > any comments from others ? ** these are two distinct timers, two distinct events (KeepAlive 11 and Hold 10) so keep the existing text > > 12 - text changed, potential consensus ** Open Delay timer in any new text please > 13a - text changed, potential consensus > 13b - no change, potential consensus **I found and find ** Transport Connection Indication & valid remote peer **verbose and clumsy to read and write and suggested/suggest ** 13 Transport connection valid indication ** 14 Transport connection invalid indication **at most, even cutting the names down to three words if possible. ** There remains plenty of scope for verbosity in the textual description although I think we should beware of nailing it down too tightly. eg Invalid in a security conscious system might come to mean this address has been added to a black list on account of its recent malicious behaviour, perhaps dynamically blocking some source port numbers on account of denial of service attacks. While valid in a lax implementation could mean anything goes, even port 179. I am ok with the existing description. > > > 14 - additional comments, no consensus > Note: No consensus on one part only > 15 - Consensus, author's choice. > 16 - Agree to your comments. > Have clarified original comment, > proposed some text. ** Open Delay timer in any new text please > 17 - consensus ** Open Delay timer in any new text please > 18 - text revised, potential consensus > 19 - no change, potential consensus > 20 - text changed, potential consensus > > >> Here's the summary of 11-20 >> >> comment: >> >> 11 - no consensus, no text supplied >> >> note: usage varies from my understanding >> of common usage. >> >> 12 - consensus, no text supplied >> see comment 9 and proposed text below. >> >> 13a - no consensus, see alternate text >> 13b - no consensus, implementation specific detail >> >> 14 - no consensus, disagree with comment >> policy issues, outside scope of document >> >> 15 - no consensus, see alternative text. >> 16 - no consensus, see my proposed text. >> >> You did not propose text >> See alternate text for events 13-17 >> Also note, that I consider that >> events 13 and 14 should go to >> optional. >> >> 17 - consensus, agree to change in text. >> 18 - agree needs to be fixed, see my alternative text. >> 19 - further clarity needed, >> please respond to question >> >> 20 - agree, text will be changed >> >> >> Sue >> >> ===================== >> >> 11) Event 10 (Hold timer) >> >> We use the hold timer for two purposes >> >> * Waiting for an open message (with a default value of 240 seconds) >> * Waiting for Keepalives (with a default value of 90 seconds) >> >> Can we use two different timers (or at least call them two different >> timer events) ? >> >> Siva: I do not believe that most implementatios >> utilize two timers. Can you indicate implementations >> that do? > > Implementations can certainly use a single timer, since at any time > at most only one of these two timers will need to be running. > > The reason for the comment is that conceptually these are two > different timers: > > One is a standard (or configurable) value, and is used to wait for an > Open Message. > > The other is negotiated, and is used to wait for Keepalives. > > I feel that representing these as two different events will make the > state machine look better. Of course, nothing prevents an implementation > from achieving this by using the same timer, or any other means as long > the same effect is realized; so existing implementations will not be > affected in any fashion. > > I am not very particular on this issue, since things will work fine > with the current means as well, do others have any comments on > modeling this as two separate events in the state machine ? > >> >> ============================================================== >> ========= >> 12) Event 12 >> >> > Event12: DelayBGP Open timer expires [changed] >> > >> > Definition: A timer that delays sending of the BGP >> > Open message for n seconds after the >> > Transport connection has been completed.= >> >> >The Definition seems to better define what the timer is, >> rather than = >> > what >> > its expiry means. Can we change it to reflect this ? >> >> new text: see comment on 9 on Timer definition, >> and earlier change. >> >> Event 12: An event generated when the Delay BGP Open timer expires. >> ========================================= >> > > Ok. > Can we add any text on the need or use of a Delay Open ? > > >> 13) Event 13, 14 definition >> >> > Event13: Transport Connection Indication & valid remote >> peer[changed] >> > >> > Definition: Event indicating that TCP connection >> > request with a valid source IP address and TCP >> > port, and valid destination IP address >> > and TCP Port. The definition of >> > invalid Source, and invalid Destination= >> > IP address is left to the implementation. >> >> >> a) How about changing this slighty to: >> % Definition: Event indicating receiving a TCP connection >> % request with a valid source IP address and TCP >> % port, and valid destination IP address= >> % and TCP Port. The definition of invalid >> My completion of your sentence: >> source, and invalid destination >> IP address >> is left to the implementation. >> >> How about: >> Definition: Event indicating the reception of a TCP connection >> request with a valid source IP address and TCP >> port, and valid destination IP address and >> TCP Port. The definition of invalid >> source address >> and port and invalid destination address >> is left to the implementation. >> >> >> Please comment if you agree with this change. > > This looks good. > >> >> b) Is it correct that Destination IP address validation is >> used when we >> do not wish to accept BGP connections on one of the local IP addresses >> (either because we wish to accept such addresses on only our loopback >> interfaces, or because the address to which we received the connection >> was not the optimal interface with which to speak to the sender) ? If >> so, can we add this to the description ? >> >> >> No, this is an implementation dependent change. > > Ok. Again, if there are any validations that are considered > desirable, adding this to the draft text as a suggestion to > implementers would be useful. > > We can leave things as is if you feel that such details should > be left out of the draft. > > >> ============================================================== >> =================== > > >> >> 14) Event 13,14 >> >> > Event13: Transport Connection Indication & valid remote >> peer[changed] >> > >> > Definition: Event indicating that transport >> > Request valid source IP address and TCP=20 >> > port, and valid destination IP address=20 >> > and TCP Port. The definition of=20 >> > invalid Source, and invalid Destination=20 >> > IP address is left to the implementation. >> > >> > a) I am unclear as to what validation of source and >> destination port = >> > mean. >> > >> > The source port's value should be of no concern to connection = >> > receipient system. >> >> The definition of "valid" is an implementation policy issue. >> Validating is the process of checking to make sure it is valid >> for the implementation's policy. > > Ok. > >> >> > The destination port cannot be invalid, since BGP would not >> receive the >> > connection request in such a situation. >> >> Again, this is a policy issue. We are providing hooks for >> policy to be engaged. Policy is outside the scope of this >> document. > > Isn't this more than just a policy issue ? It seems difficult > to identify a connection destined to a port other than to > the one on which BGP is running as an error event for BGP. > So, I don't understand how a policy can be applied in this > regard. > >> >> > b) I think it will be useful to state what we consider to >> be an invalid >> > IP address. >> >> Again, it is policy issue which is implementation dependent. > > Ok. > >> >> ==================== >> 15) Event 17 >> >> > Event17: TCP connection fails >> >> > This event can occur even when the transport connection is closed by >> > the other end. Since this does not reflect a 'failure ', >> can we change >> > the event name to >> > >> > % Event17: TCP connection termination >> >> Discussion: It both terminates from the remote site >> and can "timeout" - fail. >> >> Suggestions? I can use "disconnect", what do you think. > > That looks fine too. I don't know what the best words > to describe this are. Also, the description for this > event does a good job of explaining this, so either > the original text (connection fails) or your suggestion > (disconnect) can be used. > This really should not have been raised as a comment. :-) > >> ========================== >> >> 16) Events related to BGP Messages >> >> Can we make the definition style across these events consistent ? >> >> thanks for the comment. >> proposed text below, please agree on it. > > Yes, the text below looks fine. > > I should have phrased the original comment much better, I was > referring to consistency across the following: Events 18, 19, > 20, 21 vs. Events 23, 24, 25, 26, 27. > > This is a very minor issue, so not making these changes is fine > too. The way the latter events are described > are different from the former. I suggest adopting the style for > Events 23-27 for Events 18-21. > > As per this, the following changes would have to be made: > >> Event18: BGPOpen >> >> Definition: An event indicating that a valid Open >> message has been received. > > with > >% Event18: BGPOpen >% >% Definition: An event is generated when a valid Open >% message has been received. > >> Event19: BGPOpen with BGP Delay Open Timer running >> >> Definition: An event indicating that a valid Open >> message has been successful >> established for a peer that is >> currently delaying the sending of an >> BGP Open message. > > with > >% Event19: BGPOpen with BGP Delay Open Timer running >% >% Definition: An event is generated when a valid Open >% message has been received for a peer that >% is currently delaying the sending of a >% BGP Open message. > >> Event20: BGPHeaderErr >> >> Definition: BGP message header is not valid. > > with > >% Event20: BGPHeaderErr >% >% Definition: An event is generated when a received BGP >% message header is not valid. > >> Event21: BGPOpenMsgErr >> >> Definition: An BGP Open message has been received >> with errors. > > with > >% Event21: BGPOpenMsgErr >% >% Definition: An event is generated when BGP Open message >% with errors has been received. > > >> >> proposed text replacement >> ------------------------ >> Event13: TCP connection indication and valid remote peer >> >> Definition: Event indicating the local system reception of >> a TCP connection request with a valid source >> IP address and TCP port, and valid destination >> IP address and TCP Port. The definition of >> invalid source, and invalid destination >> IP address is left to the implementation. >> >> BGP's destination port SHOULD be port >> 179 as defined by IANA. >> >> TCP connection request is denoted by >> the local system receiving a TCP SYN. >> >> Status: Mandatory (Optional) >> >> >> Event14: RCV TCP connection indication with invalid source or >> destination >> >> Definition: Event indicating the local system reception of >> a TCP connection request with either >> an invalid source address or port >> number or an invalid destination >> address or port number. >> >> BGP destination port number SHOULD be 179 >> as defined by IANA. >> >> Again, a TCP connection request >> denoted by local system receiving a TCP >> SYN. >> >> Status: Mandatory (Optional) >> >> Event15: TCP connection request sent received an ACK. >> >> Definition: Event indicating the Local system's request >> to establish a TCP connection to the remote >> peer. >> >> The local system's TCP session sent a TCP >> SYN, and received a TCP SYN, ACK pair of >> messages, and Sent a TCP ACK. >> >> Status: Mandatory >> >> Event16: TCP connection confirmed >> >> Definition: Event indicates that the local system receiving >> a confirmation that the TCP connection has >> been established by the remote site. >> >> The remote peer's TCP engine sent a TCP SYN. >> The local peer's TCP engine sent a SYN, ACK >> pair, and now has received a final ACK. >> >> Status: Mandatory >> >> Event17: TCP connection fails >> >> Definition: Event indicates that the local system has >> received a TCP connection failure notice. >> >> The remote BGP peer's TCP machine could have >> sent a FIN. The local peer would respond >> with a FIN-ACK. Another alternative is that >> the local peer indicated a timeout in the >> TCP session and downed the connection. >> >> Status: Mandatory >> >> --------------------- >> >> >> 17) Event 19 Definition >> >> > Event19: BGPOpen with BGP Delay Open Timer running [changed] >> > >> > Definition: An event indicating that a valid Open=20 >> > Message has been successful=20 >> > established for a peer that is=20 >> > currently delaying the sending of an=20 >> > BGP Open message.=20 >> >> >> The definition appears a little fragmented. How about: >> >> % Definition: An event indicating that a valid OPEN >> % Message has been received for a peer >> % that has a successfully established >> % transport connection and is currently >> % delaying the seending of a BGP open >> % message >> >> >> agreed, changed in text. I will send out definition >> text for review in a separate message. > > Ok. > >> ============================================================== >> 18) Event 22 Definition >> >> a)=20 >> >> > Event22: Open collision discard >> > >> > Definition: An event generated administratively=20 >> > when a connection Collision has been=20 >> > detected while processing an incoming =20 >> > Open message. This connection has been=20 >> >> Isn't this event 'automatically' generated, since it is a system >> generated event ?=20 >> >> response: How this generated is implementation specific. The >> "administratively" is to cover policy. >> >> b) >> >> > Event 22 is an administrative could=20 >> > occur if FSM is implemented as two=20 >> > >> >> The word event is missing. How about >> >> % Event 22 is an automatic event that >> % could occur if FSM is implemented as two=20 >> >> >> I've re-written the whole definition. What do you think about: >> >> Event22: Open collision dump >> >> Definition: An event generated administratively >> when a connection collision has been >> detected while processing an incoming >> OPEN message and this connection has been >> selected to disconnected. See Section >> 6.8 for more information on collision >> detection. >> >> Event22 is an administrative based only >> implementation specific policy. This >> Event may occur if the FSM is implemented >> as two linked state machines. >> >> Please confirm that you like this. > > Agreed. > >> >> ==================================== >> 19) FSM Description: ConnectRetry Count >> >> > The Connect Retry count is updated by the FSM but never used. In the >> > absence of peer oscillation damping, will this be used to stop >> > connection establishment attempts after a certain maximum number ? >> >> Yes, this is either implementation specific or >> is it based on the peer oscillation damping draft. >> >> > Can we include the use of this counter in some place ? >> >> Connect retry counter >> 1) Will be utilized by the peer oscillation >> damping drat. >> 2) Will be included in bgp-4-mibv2-xx. >> I just check and I didn't find it. >> >> Do you still want text in the main > > We can leave the main text as is. > >> >> ==================================== >> >> 20) Handling Event 7 (Auto Stop) to Idle State processing: >> >> The handling of Event 7 is missing from the Idle State processing. Can >> we add this ? How about replacing >> >> > An manual stop event (Event2) is ignored in the Idle state. >> >> with >> >> % Manual stop (Event 2) and Auto stop (Event 7) events are ignored >> % in the Idle state >> >> we'll add that in the text. > > Ok. > >> >> ============================== >> > >Siva >(siva@ctd.hcltech.com) Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id NAA06373 for <idr-archive@nic.merit.edu>; Tue, 10 Dec 2002 13:09:36 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 15530912A9; Tue, 10 Dec 2002 13:09:08 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id C255C912AB; Tue, 10 Dec 2002 13:09:07 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 9C96C912A9 for <idr@trapdoor.merit.edu>; Tue, 10 Dec 2002 13:09:05 -0500 (EST) Received: by segue.merit.edu (Postfix) id 820BB5DEF5; Tue, 10 Dec 2002 13:09:05 -0500 (EST) Delivered-To: idr@merit.edu Received: from ganesh.ctd.hctech.com (unknown [202.54.64.2]) by segue.merit.edu (Postfix) with ESMTP id E75575DEE8 for <idr@merit.edu>; Tue, 10 Dec 2002 13:09:01 -0500 (EST) Received: by GANESH with Internet Mail Service (5.5.2653.19) id <YT9SJPJY>; Tue, 10 Dec 2002 23:48:30 +0530 Message-ID: <55E277B99171E041ABF5F4B1C6DDCA06EA018C@haritha.hclt.com> From: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com> To: Susan Hares <shares@nexthop.com>, Yakov Rekhter <yakov@juniper.net>, Jeffrey Haas <jhaas@nexthop.com>, idr@merit.edu Cc: "Venu Kumar G - CTD, Chennai." <venug@ctd.hcltech.com> Subject: RE: Comments 11-20 Date: Tue, 10 Dec 2002 23:41:06 +0530 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-idr@merit.edu Precedence: bulk Hello, Revised summary: 11 - No consensus, but current text can stand as is any comments from others ? 12 - text changed, potential consensus 13a - text changed, potential consensus 13b - no change, potential consensus 14 - additional comments, no consensus Note: No consensus on one part only 15 - Consensus, author's choice. 16 - Agree to your comments. Have clarified original comment, proposed some text. 17 - consensus 18 - text revised, potential consensus 19 - no change, potential consensus 20 - text changed, potential consensus > Here's the summary of 11-20 > > comment: > > 11 - no consensus, no text supplied > > note: usage varies from my understanding > of common usage. > > 12 - consensus, no text supplied > see comment 9 and proposed text below. > > 13a - no consensus, see alternate text > 13b - no consensus, implementation specific detail > > 14 - no consensus, disagree with comment > policy issues, outside scope of document > > 15 - no consensus, see alternative text. > 16 - no consensus, see my proposed text. > > You did not propose text > See alternate text for events 13-17 > Also note, that I consider that > events 13 and 14 should go to > optional. > > 17 - consensus, agree to change in text. > 18 - agree needs to be fixed, see my alternative text. > 19 - further clarity needed, > please respond to question > > 20 - agree, text will be changed > > > Sue > > ===================== > > 11) Event 10 (Hold timer) > > We use the hold timer for two purposes > > * Waiting for an open message (with a default value of 240 seconds) > * Waiting for Keepalives (with a default value of 90 seconds) > > Can we use two different timers (or at least call them two different > timer events) ? > > Siva: I do not believe that most implementatios > utilize two timers. Can you indicate implementations > that do? Implementations can certainly use a single timer, since at any time at most only one of these two timers will need to be running. The reason for the comment is that conceptually these are two different timers: One is a standard (or configurable) value, and is used to wait for an Open Message. The other is negotiated, and is used to wait for Keepalives. I feel that representing these as two different events will make the state machine look better. Of course, nothing prevents an implementation from achieving this by using the same timer, or any other means as long the same effect is realized; so existing implementations will not be affected in any fashion. I am not very particular on this issue, since things will work fine with the current means as well, do others have any comments on modeling this as two separate events in the state machine ? > > ============================================================== > ========= > 12) Event 12 > > > Event12: DelayBGP Open timer expires [changed] > > > > Definition: A timer that delays sending of the BGP > > Open message for n seconds after the > > Transport connection has been completed.= > > >The Definition seems to better define what the timer is, > rather than = > > what > > its expiry means. Can we change it to reflect this ? > > new text: see comment on 9 on Timer definition, > and earlier change. > > Event 12: An event generated when the Delay BGP Open timer expires. > ========================================= > Ok. Can we add any text on the need or use of a Delay Open ? > 13) Event 13, 14 definition > > > Event13: Transport Connection Indication & valid remote > peer[changed] > > > > Definition: Event indicating that TCP connection > > request with a valid source IP address and TCP > > port, and valid destination IP address > > and TCP Port. The definition of > > invalid Source, and invalid Destination= > > IP address is left to the implementation. > > > a) How about changing this slighty to: > % Definition: Event indicating receiving a TCP connection > % request with a valid source IP address and TCP > % port, and valid destination IP address= > % and TCP Port. The definition of invalid > My completion of your sentence: > source, and invalid destination > IP address > is left to the implementation. > > How about: > Definition: Event indicating the reception of a TCP connection > request with a valid source IP address and TCP > port, and valid destination IP address and > TCP Port. The definition of invalid > source address > and port and invalid destination address > is left to the implementation. > > > Please comment if you agree with this change. This looks good. > > b) Is it correct that Destination IP address validation is > used when we > do not wish to accept BGP connections on one of the local IP addresses > (either because we wish to accept such addresses on only our loopback > interfaces, or because the address to which we received the connection > was not the optimal interface with which to speak to the sender) ? If > so, can we add this to the description ? > > > No, this is an implementation dependent change. Ok. Again, if there are any validations that are considered desirable, adding this to the draft text as a suggestion to implementers would be useful. We can leave things as is if you feel that such details should be left out of the draft. > ============================================================== > =================== > > 14) Event 13,14 > > > Event13: Transport Connection Indication & valid remote > peer[changed] > > > > Definition: Event indicating that transport > > Request valid source IP address and TCP=20 > > port, and valid destination IP address=20 > > and TCP Port. The definition of=20 > > invalid Source, and invalid Destination=20 > > IP address is left to the implementation. > > > > a) I am unclear as to what validation of source and > destination port = > > mean. > > > > The source port's value should be of no concern to connection = > > receipient system. > > The definition of "valid" is an implementation policy issue. > Validating is the process of checking to make sure it is valid > for the implementation's policy. Ok. > > > The destination port cannot be invalid, since BGP would not > receive the > > connection request in such a situation. > > Again, this is a policy issue. We are providing hooks for > policy to be engaged. Policy is outside the scope of this > document. Isn't this more than just a policy issue ? It seems difficult to identify a connection destined to a port other than to the one on which BGP is running as an error event for BGP. So, I don't understand how a policy can be applied in this regard. > > > b) I think it will be useful to state what we consider to > be an invalid > > IP address. > > Again, it is policy issue which is implementation dependent. Ok. > > ==================== > 15) Event 17 > > > Event17: TCP connection fails > > > This event can occur even when the transport connection is closed by > > the other end. Since this does not reflect a 'failure ', > can we change > > the event name to > > > > % Event17: TCP connection termination > > Discussion: It both terminates from the remote site > and can "timeout" - fail. > > Suggestions? I can use "disconnect", what do you think. That looks fine too. I don't know what the best words to describe this are. Also, the description for this event does a good job of explaining this, so either the original text (connection fails) or your suggestion (disconnect) can be used. This really should not have been raised as a comment. :-) > > > ========================== > > 16) Events related to BGP Messages > > Can we make the definition style across these events consistent ? > > thanks for the comment. > proposed text below, please agree on it. Yes, the text below looks fine. I should have phrased the original comment much better, I was referring to consistency across the following: Events 18, 19, 20, 21 vs. Events 23, 24, 25, 26, 27. This is a very minor issue, so not making these changes is fine too. The way the latter events are described are different from the former. I suggest adopting the style for Events 23-27 for Events 18-21. As per this, the following changes would have to be made: > Event18: BGPOpen > > Definition: An event indicating that a valid Open > message has been received. with % Event18: BGPOpen % % Definition: An event is generated when a valid Open % message has been received. > Event19: BGPOpen with BGP Delay Open Timer running > > Definition: An event indicating that a valid Open > message has been successful > established for a peer that is > currently delaying the sending of an > BGP Open message. with % Event19: BGPOpen with BGP Delay Open Timer running % % Definition: An event is generated when a valid Open % message has been received for a peer that % is currently delaying the sending of a % BGP Open message. > Event20: BGPHeaderErr > > Definition: BGP message header is not valid. with % Event20: BGPHeaderErr % % Definition: An event is generated when a received BGP % message header is not valid. > Event21: BGPOpenMsgErr > > Definition: An BGP Open message has been received > with errors. with % Event21: BGPOpenMsgErr % % Definition: An event is generated when BGP Open message % with errors has been received. > > proposed text replacement > ------------------------ > Event13: TCP connection indication and valid remote peer > > Definition: Event indicating the local system reception of > a TCP connection request with a valid source > IP address and TCP port, and valid destination > IP address and TCP Port. The definition of > invalid source, and invalid destination > IP address is left to the implementation. > > BGP's destination port SHOULD be port > 179 as defined by IANA. > > TCP connection request is denoted by > the local system receiving a TCP SYN. > > Status: Mandatory (Optional) > > > Event14: RCV TCP connection indication with invalid source or > destination > > Definition: Event indicating the local system reception of > a TCP connection request with either > an invalid source address or port > number or an invalid destination > address or port number. > > BGP destination port number SHOULD be 179 > as defined by IANA. > > Again, a TCP connection request > denoted by local system receiving a TCP > SYN. > > Status: Mandatory (Optional) > > Event15: TCP connection request sent received an ACK. > > Definition: Event indicating the Local system's request > to establish a TCP connection to the remote > peer. > > The local system's TCP session sent a TCP > SYN, and received a TCP SYN, ACK pair of > messages, and Sent a TCP ACK. > > Status: Mandatory > > Event16: TCP connection confirmed > > Definition: Event indicates that the local system receiving > a confirmation that the TCP connection has > been established by the remote site. > > The remote peer's TCP engine sent a TCP SYN. > The local peer's TCP engine sent a SYN, ACK > pair, and now has received a final ACK. > > Status: Mandatory > > Event17: TCP connection fails > > Definition: Event indicates that the local system has > received a TCP connection failure notice. > > The remote BGP peer's TCP machine could have > sent a FIN. The local peer would respond > with a FIN-ACK. Another alternative is that > the local peer indicated a timeout in the > TCP session and downed the connection. > > Status: Mandatory > > --------------------- > > > 17) Event 19 Definition > > > Event19: BGPOpen with BGP Delay Open Timer running [changed] > > > > Definition: An event indicating that a valid Open=20 > > Message has been successful=20 > > established for a peer that is=20 > > currently delaying the sending of an=20 > > BGP Open message.=20 > > > The definition appears a little fragmented. How about: > > % Definition: An event indicating that a valid OPEN > % Message has been received for a peer > % that has a successfully established > % transport connection and is currently > % delaying the seending of a BGP open > % message > > > agreed, changed in text. I will send out definition > text for review in a separate message. Ok. > ============================================================== > 18) Event 22 Definition > > a)=20 > > > Event22: Open collision discard > > > > Definition: An event generated administratively=20 > > when a connection Collision has been=20 > > detected while processing an incoming =20 > > Open message. This connection has been=20 > > Isn't this event 'automatically' generated, since it is a system > generated event ?=20 > > response: How this generated is implementation specific. The > "administratively" is to cover policy. > > b) > > > Event 22 is an administrative could=20 > > occur if FSM is implemented as two=20 > > > > The word event is missing. How about > > % Event 22 is an automatic event that > % could occur if FSM is implemented as two=20 > > > I've re-written the whole definition. What do you think about: > > Event22: Open collision dump > > Definition: An event generated administratively > when a connection collision has been > detected while processing an incoming > OPEN message and this connection has been > selected to disconnected. See Section > 6.8 for more information on collision > detection. > > Event22 is an administrative based only > implementation specific policy. This > Event may occur if the FSM is implemented > as two linked state machines. > > Please confirm that you like this. Agreed. > > ==================================== > 19) FSM Description: ConnectRetry Count > > > The Connect Retry count is updated by the FSM but never used. In the > > absence of peer oscillation damping, will this be used to stop > > connection establishment attempts after a certain maximum number ? > > Yes, this is either implementation specific or > is it based on the peer oscillation damping draft. > > > Can we include the use of this counter in some place ? > > Connect retry counter > 1) Will be utilized by the peer oscillation > damping drat. > 2) Will be included in bgp-4-mibv2-xx. > I just check and I didn't find it. > > Do you still want text in the main We can leave the main text as is. > > ==================================== > > 20) Handling Event 7 (Auto Stop) to Idle State processing: > > The handling of Event 7 is missing from the Idle State processing. Can > we add this ? How about replacing > > > An manual stop event (Event2) is ignored in the Idle state. > > with > > % Manual stop (Event 2) and Auto stop (Event 7) events are ignored > % in the Idle state > > we'll add that in the text. Ok. > > ============================== > Siva (siva@ctd.hcltech.com) Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id QAA00784 for <idr-archive@nic.merit.edu>; Mon, 9 Dec 2002 16:50:38 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id EF68D91274; Mon, 9 Dec 2002 16:50:11 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id B2E1B91275; Mon, 9 Dec 2002 16:50:10 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 373A391274 for <idr@trapdoor.merit.edu>; Mon, 9 Dec 2002 16:50:09 -0500 (EST) Received: by segue.merit.edu (Postfix) id 1834A5E075; Mon, 9 Dec 2002 16:50:09 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 7AE135DE1F for <idr@merit.edu>; Mon, 9 Dec 2002 16:50:08 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB9Lo6M54518 for idr@merit.edu; Mon, 9 Dec 2002 16:50:06 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB9LnoC54464 for <idr@merit.edu>; Mon, 9 Dec 2002 16:49:50 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Subject: RE: damping persistent peer oscillations X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 Date: Mon, 9 Dec 2002 16:49:50 -0500 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA616A5@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: damping persistent peer oscillations Thread-Index: AcKdavRo3W9hPEyyRMqO6rtQrK+uYQCYaH9w From: "Susan Hares" <shares@nexthop.com> To: "Tom Petch" <nwnetworks@dial.pipex.com>, "Natale, Jonathan" <JNatale@celoxnetworks.com>, "idr" <idr@merit.edu> Cc: <yakov@juniper.net> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id QAA00784 Tom: peer oscillation damping ---- The document -19 draft already has the corrections (thanks to yakov). thanks for catching this. I'll double check prior to release of the document. Sue -----Original Message----- From: Tom Petch [mailto:nwnetworks@dial.pipex.com] Sent: Friday, December 06, 2002 4:02 PM To: Susan Hares; Natale, Jonathan; idr Subject: Re: damping persistent peer oscillations ** comment inline -----Original Message----- From: Susan Hares <shares@nexthop.com> To: Natale, Jonathan <JNatale@celoxnetworks.com>; idr@merit.edu <idr@merit.edu> Date: 06 December 2002 19:18 Subject: RE: damping persistent peer oscillations Jonathan: The draft is below. I'll be updating the draft-hares-bgp-backoff-00.txt draft this month. Please send comments. The base specification will simply state "damping persistent BGP adjacency flapping ** aka peer oscillation damping :-) are outside the scope of this document. Sue --------------------- Internet Draft S. Hares Document: draft-hares-bgp-backoff-00.txt NextHop Technologies, Inc. Expires: December 2002 June 2002 BGP Peer Restart Backoff Mechanisms -----Original Message----- From: Natale, Jonathan [mailto:JNatale@celoxnetworks.com] Sent: Friday, December 06, 2002 10:47 AM To: idr@merit.edu Subject: damping persistent peer oscillations >From Draft 18: "Local system automatically starts the BGP peer connection with persistent peer oscillation damping enabled. The exact method of damping persistent peer oscillations is left up to the implementation. These methods of damping persistent BGP adjacency flapping are outside the scope of this document." -- is there a draft describing a "method of damping persistent peer oscillations" ? Tom Petch nwnetworks@dial.pipex.com Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id QAA00653 for <idr-archive@nic.merit.edu>; Mon, 9 Dec 2002 16:46:36 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 35D0C91233; Mon, 9 Dec 2002 16:45:21 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id E9AB191274; Mon, 9 Dec 2002 16:45:20 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 7C76891233 for <idr@trapdoor.merit.edu>; Mon, 9 Dec 2002 16:45:18 -0500 (EST) Received: by segue.merit.edu (Postfix) id F414A5E07A; Mon, 9 Dec 2002 16:45:16 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id E86DE5E075 for <idr@merit.edu>; Mon, 9 Dec 2002 16:45:14 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB9LjDn54334 for idr@merit.edu; Mon, 9 Dec 2002 16:45:13 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB9Lj2C54309 for <idr@merit.edu>; Mon, 9 Dec 2002 16:45:02 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Subject: RE: Response to FSM input - Comments 1-10 X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 Date: Mon, 9 Dec 2002 16:45:02 -0500 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA616A4@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: Response to FSM input - Comments 1-10 Thread-Index: AcKdFw/Ma0ArdhU1STmrF5x92p3z/QARRo5g From: "Susan Hares" <shares@nexthop.com> To: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com>, "Yakov Rekhter" <yakov@juniper.net>, "Jeffrey Haas" <jhaas@nexthop.com> Cc: <idr@merit.edu>, "Venu Kumar G - CTD, Chennai." <venug@ctd.hcltech.com> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id QAA00653 Siva: Thanks for such a quick turn around. So.. our summary is 1 - consensus, no text change 2 - no consensus, see text change 3 - mostly consensus, Policy in implementations may dictate some options per peer and some options per implementation. I suggest we specify the list you came up with plus Idle Hold Timer in 8.0. 4 - consensus, text change [done] 5 - no consensus, see comment below 5a - answer question [no text intended] 5b.1 - text proposed 5b.2 - your text or mine? 6 - consensus on text, answer given below 7 - no consensus, text proposal below 2 of the 3 events could be considered infeasible. The 3rd event all add. Please review the text. 8 - consensus, no text change 9 - consensus, text change [done] 10 - consensus, text changed [done] Note 9 + 10 are taken together Comments to resolve: 2,3, 5, 7, Thanks for the quick response. Sue Hares ===================== Comment 2 - Event8: Idle hold timer expires Definition: An Event generated when the Idle Hold Timer expires. The Idle Hold Timer is only used when the persistent peer oscillation damping function is enabled. % % Implementations not implemented persistent % peer oscillations damping functions may not % have the Idle Hold Timer. Is this approach ok? ============ Comment 3 The following will be added to 8.0 Optional parameters that may be supported either per connection or per implementation: 1) Delay Open flag 2) Delay Open Timer 3) Perform automatic start flag 4) Passive TCP establishment flag 5) BGP stop_peer_flag flag %6) Idle Hold timer 7) Perform automatic stop flag 8) Perform Collision detect in Establish mode flag ===================== Comment 4: consensus, change made to Event 1 and 2 =============== Comment 5: 5a: Automatic stop and stop are implemented by cisco's policy engine. The origination of the damping peer oscillation was in result to a bug that was listed in comment 5. 5b: 1.) Example of "automatic disconnect" in Established state Is the text you wanted moved: An example automatic stop event is exceeding the number of prefixes for a given peer and the local system automatically disconnecting the peer. If so, I can move this to the automatic stop event 2.) Text describing the optional parameters. We can add text for the optional parameters to describe their function. Can you propose some text? ====================== Comment 6 > I am ok to leave the text as is, based on text in 8.2.1.1. However, > I do not quite understand your point. In the case of an un-configured > peer, the local system cannot 'establish a connection', so is the > unconfigured peer case still valid in this context ? Unconfigured + passive TCP connection + Delay Open is valid. ============================================ Comment 7 > > > We already have > Event 3 - Automatic Start > Event 5 - Automatic start with bgp_stop_flap option set > > > To make things consistent, shouldn't we either > > a) Add 3 new events : 1) Manual start with bgp_stop flap option set > 2) Manual start with passive TCP establishment > and bgp_stop_flap option set > 3) Automatic start with passive TCP establishment > and bgp_stop_flap option set I like option A, it is clearer. The Manual start or stop terminates the bgp peer oscillation damping. The theory is the operator "did" something, therefore the events need to chagne. Since these events do not have valid implementation in the peer flap damping, I did not include them. #3 is something I missed. I will add that event. Is that resolution OK, or I can put all 3 new events in. ========================= -----Original Message----- From: Sivananda Ramnath - CTD, Chennai. [mailto:siva@ctd.hcltech.com] Sent: Friday, December 06, 2002 6:05 AM To: Susan Hares; Yakov Rekhter; Jeffrey Haas Cc: idr@merit.edu; Sivananda Ramnath (E-mail); Venu Kumar G - CTD, Chennai. Subject: RE: Response to FSM input - Comments 1-10 Hello, Revised summary: Comment 1 - No text change, Potential consensus Comment 2a - No consensus Comment 2b - No consensus Comment 3 - New text suggested, no consensus Comment 4 - Resolved, text changed. Can we consider this closed ? Comment 5 - No text change, Potential consensus Comment 6 - No consensus Comment 7 - No consensus, comment re-phrased. Comment 8 - No text change, potential consensus Comment 9 - Resolved, text changed. Can we consider this closed ? Comment 10 - Resolved, text changed. Can we consider this closed ? > > Comment 1 - no consensus > Comment 2a - out of date (No IdleHold Timer exists) > Comment 2b - no consensus, can't remove due to OpenSent state > action. > > Comment 3 - no text proposed, do we want all options > specified in 8.1 (7 options) > Please review and comment on text > > Comment 4 - agree > Comment 5 - no text proposed, no consensus > Comment 6 - no consensus on the text > Comment 7 - see comment 3, no consensus > Comment 8 - question, no text > Comment 9 - agree, I supplied text > Comment 10 - agree to remove text in option, > different text for definition. > > > Sue Hares > > ========== > > comment 1: > > 1) removal of peer oscillation damping > > The reason we should not remove the peer oscillation damping > from the state machine: > > 1) Deployed implementations support peer oscillation damping > 2) Hooks for the additions in the FSM cannot be added later. > > These hooks are optional and do not need to be implemented. I understand. I am not trying to object to peer oscillation damping, I think it is a good idea and we have included it in our implementation as well. I was suggesting that instead of a partial description in this draft, it be completely described in the draft on peer oscillation damping. However, I do see your point, and unless there are any objections from others, I think we have consensus on this issue. > > > 2) Idle HoldTimer in draft-18 > > Perhaps this is an old comment. I don't find IdleHoldTimer > in the messages. It is listed as Event 8. Perhaps this is what needs to be removed ? > > 3) HoldTime and HoldTimer > > The Hold Time is specified on page 13 of the draft-18 specification > as to the value and size. The text in OpenSent requires > both a HoldTime and HoldTimer value. > =========== > OpenSent: > > In this state BGP waits for an Open Message from its peer. [SNIP] > > =================== I understand. However, on this basis, we would also need a Keepalive time (since the basis on wihch this is be arrived at can be configurable). I don't know whether we do need to make such changes to the draft, and I am OK to leaving this portion as it stands. > The earlier comments on the list indicated we need to collect what > timers or Timer values were utilized in the FSM. I believe Alex > made this comment last January. (the list archives will have > this information.) > > So, I don't see any way to The DelayBgpOpenTimer is missing from the session attributes list, how about including it ? > > Comment 3: optional parameters specified > > Please suggest text. We have the following optional parameters: > > 1) Delay Open > 2) Accept unconfigured peers > 3) automatic start [Event 3] > 4) passive TCP establishment (optional) > 4 + 5 = Event 5 > 5) BGP_stop_flap (BGP Peer oscillation) [Event 6] > 6)automatic stop [Event 7] > 7)Collision detect in Established State > > If you want some of the optional parameters specified, let's > catch all. Please suggest text and location. It could > go in 8.1. My view is this can be split into two categories: Optional attributes exist per session, and optional attributes for the entire system. How about adding the following to section 8 (immediately before 8.1) % Optional session attributes that a implementation may support % for a session are: % % 1) DelayOpen flag % 2) DelayOpen Timer % 3) Perform automatic start % 4) Passive TCP establishment % 5) Perform automatic stop % 6) BGP Perform % % % Implementations may also support the following optinal attributes % that modify the behavior of the state machine % % 1) Accept un-configured peers % 2) Perform collision detect in Established state I guess it is possible to have some of these things listed under both categories. Am I opening up a can of worms here ? :-) > -------------------------------- > Comment 4 > > Old Text: > Event 1: Manual Start > Definition: Administrator manually starts peer connection. > > ... > Event 2: Manual Stop > Definition: Local system administrator manually stops the > peer connection. > New Text: > Event1: Manual start > Definition: Local system administrator manually starts the > bgp connection. > > ... > Event 2: Manual stop: > Definition: Local system administrator manually stops the > bgp connection. > > agree? Agreed. > --------------- > Comment 5: > > 5) Event3, Event5, Event6 and Event7 > > 5a) > ========= > Event 3/6/7 - cisco indicated the automatic stop, see the NANOG > archives for a discussion of the Bad Route > > http://www.merit.edu/mail.archives/nanog/2001-06/msg00805.html > > The "event" sequences are new with this version of the FSM > to identify particular events by number. I am unclear on what this issue is! > 5b > ====== > > Can we give examples of these events, as implemneted by curent > implementations ? This will be of help to new implementaions. > > An example of Auto Stop is already present in the Update event > handling section, and we can move that to this place as well. > > I believe that examples are outside the scope of this specification. > If you have text clarifying the event descriptions, please send it in. Well, the current text in the established state already gives an example for an automatic start event. I was suggesting that such examples be moved to the event description section. I also think adding text to clarify the intent behind optional events, and giving examples would help, especially for new implementors. However, if you feel this is outside the scope of the RFC, we can leave things as is. > > > ======================================================= > > 6) Event 4,5 Definition > > > Definition: Administrator manually start the peer > > Connection, but has the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing the connection. > > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > > >My understanding is this is used to indicate a peer configuration > >where we start the state machine, and wait for an incoming connection > >request from the configured peer. Can we re-phrase this section to > >indicate this. > > >How about something like: > > > >% The passive flag indicates that the state machine will wait for > >% specified peer to initiate a connection with the local system. If > >% this does not happen within a specific time (hold time), the local > >% system will then also attempt to initiate connection with the > >% specified peer. > > The text in 8.2.1.1 indicates the definition of the passive flag. Yes. > > 6a) > ========== > My understanding of your text is that you want to replace in both > sets of text: > > "The passive flag indicates the peer will listen prior to > establishing a connection". > > with: > > "The passive flag indicates that the state machine will wait for the > specified peer to initiate a connection with a local system. > > The problem with this sentence is that in the "unconfigured" case > the phrase "specified" peer is confusing. I think the original text > is clearer. I am ok to leave the text as is, based on text in 8.2.1.1. However, I do not quite understand your point. In the case of an un-configured peer, the local system cannot 'establish a connection', so is the unconfigured peer case still valid in this context ? > 6b) > ========== > If this does not happen within a specific time (hold time), the local > system will then also attempt to initiate (a) connection with the > specified peer. > > My comments: Again, the "specified peer" term is confusing. Also, > the 2nd half of the statement mixes the actions of the state machine > with the events. I believe this muddies the text instead of > clarifying it. Please refer to my comments for 6a. I think a reason for this is the requirement (in 8.2.1) that a FSM be established a-priori for "accepting each incoming TCP connection for which the peer has not beet identified". Is the intent to establish a FSM session for accepting TCP connections from un-configured peers ? If so, I am of the opinion that this is purely an implementation issue, and need not be a requirement in the spec. > > ================= > Comment 7 - > > Event 6 - (not specified, please confirm) > > > Won't a variant of this with bgp_stop_flap option set be requried ? > > We can also achieve the same by using the bgp_stop-Flap option as a > > flag that is provided as an input to the state machine. > > see comment number 3. bgp_stop_flap is an option. > > This state machine description has many events to describe the > mixture of options and actions. I believe it has a better clarity > than testing lots of flags within a state. > > A proof in case, is that many people are making comments about > specific transitions. I tink I should have phrased my question much better. We already have Event 3 - Automatic Start Event 5 - Automatic start with bgp_stop_flap option set To make things consistent, shouldn't we either a) Add 3 new events : 1) Manual start with bgp_stop flap option set 2) Manual start with passive TCP establishment and bgp_stop_flap option set 3) Automatic start with passive TCP establishment and bgp_stop_flap option set or b) Remove Event 6, and rely on a flag to tell us wether peer flap damping is to be performed for the session or not. > > ============ > Comment 8 - > > > 8) Event 5 Definition > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > > Is it correct that this event will occur only when we want > to restart > > a connection (after it had been terminated due to some reason beside > > administrative action) that we had accepted from an > unconfigured peer ? > > The automatic start function is an implementation specific mechanism. > This text does not seek to restrict it in any fashion. Well, for new implementors, such clarifications will help undertsand how to generate and make use of this event. Again, I am OK with leaving text as is. > > ================ > 9) Timer Events Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > > > Event9: connect retry timer expires > > > > Definition: An event triggered by the expiration of=20 > > Event10: hold time expires > > > > Definition: An event generated when the HoldTimer=20 > > We can use any one of these terms (trigerred by/generated > when) for all > timer event definitions to make it look more consistent. > > agree: Will use "generate" > > Event 8: An event generated when the Idle Hold timer expires. > Event 9: An event generated when the ConnectRetry timer expires. > Event 10: An event generated when the Hold timer expires. > Event 11: An event generated when the Keepalive timer expires > Event 12: An event generated when the Delay BGP Open timer expires. > > Please confirm that this change to the definitions is agreeable to > you. Agreed. I think we have consensus on this issue, unless anyone else has any comments on this. > > ================ > 10) Event 8 Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > Hold Timer is only used when persistent =20 > > BGP oscillation damping functions are=20 > > enabled. =20 > > > > Status: Optional. Used when persistent > > BGP peer oscillation damping functions > > are enabled.=20 > > > The status repeats the senetnce in the definition. Can we remove the > sentence from the Status, or replace with a different definition. > > Do you think the defenition below would fit ? > > % An event trigerred by the expiry of the Idle Hold timer, indicating > % that the session has completed waiting for a back-off period to > % prevent flap, and may now proceed to take any further systemm > % initiated action for this session. > > text change suggestion for 1st sentence that I'll agree to. > > An event trigerred by the expiry of the Idle Hold timer, indicating > that the session has completed waiting for a back-off period to > prevent bgp peer oscillation. > > Please confirm that you will agree to this. Agreed. I think we have consensus on this issue, unless anyone else has any comments on this. Siva (siva@ctd.hcltech.com)> Original text of comments 1-10 follow: > > ------- Message 4 > > Date: Fri, 29 Nov 2002 19:46:06 +0530 > From: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com> > To: skh@nexthop.com, Yakov Rekhter <yakov@juniper.net> > cc: skh@ndzh.com > Subject: Comments on draft 18 > > This message is in MIME format. Since your mail reader does > not understand > this format, some or all of this message may not be legible. > > - ------_=_NextPart_000_01C297B1.DB0F1700 > Content-Type: text/plain; > charset="iso-8859-1" > > Hello, > > Attached are some comments that I have on draft-ietf-idr-bgp4-18. > > Please have a look at these and tell me if the same can > be raised to the > mailing list as well. > > Thanks, > Siva > (siva@ctd.hcltech.com) > > > > - ------_=_NextPart_000_01C297B1.DB0F1700 > Content-Type: text/plain; > name="BGP_Issues.txt" > Content-Transfer-Encoding: quoted-printable > Content-Disposition: attachment; > filename="BGP_Issues.txt" > > All quoted sections of the original document have been > prefixed by a '> = > ' > All lines of proposed text have been prefixed by a '% ' > > > 1) Performing of peer oscillation damping: > > Since this feature is going to be added in a new draft, and its > addition will change the operation of the state machine, can we remove > all mention of it in the state machine ? As part of this removal, can > we also remove the IdleHold Timer from the FSM since it is not useful > in the absence of peer oscillation damping ? > > The draft that describes this procedure can then describe the change > in the state machine required to do this. > > > 2) Session Attributes: > > > Session Attributes required for each connection are;=20 > > > 1) state > > 2) Connect Retry timer > > 3) Hold timer > > 4) Hold time > > 5) Keepalive timer > > Why have a Hold Timer and a Hold Time ? Can we replace this with just > Hold timer ? > > Can we also add the following session attributes: > > a) DelayBgpOpenTimer > b) IdleHold Timer (in case we choose not to remove this from the bas = > FSM) > > Can we also add the following flag to the session attributes: > a) DelayOpen Flag > > > 3) All sessions attributes: > > There are some attributes that are common to all sessions of the = > system. > Can we call these out as well: > > * Accept Connections from unconfigured peers (Enabled/Disabled) > * Peer Oscillation Dampening (Enabled/Disabled) (In case we choose > not to remove it from base spec) > > > 4) Description for Event1/Event2 > > > Event1: Manual start=20 > > > > Definition: Administrator manually starts peer > > connection > > > Event2: Manual stop > > > Definition: Local system administrator manually=20 > > stops the peer connection. > > We can make these look consistent by using the Term Administrator or > Local System administrator in both cases. > > > 5) Event3, Event5, Event6 and Event7 > > Can we give examples of these events, as implemneted by curent > implementations ? This will be of help to new implementaions. > > An example of Auto Stop is already present in the Update event > handling section, and we can move that to this place as well. > > > 6) Event 4,5 Definition > > > Definition: Administrator manually start the peer > > Connection, but has the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing the connection. > > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > > My understanding is this is used to indicate a peer configuration > where we start the state machine, and wait for an incoming connection > request from the configured peer. Can we re-phrase this section to > indicate this. > > How about something like: > > % The passive flag indicates that the state machine will wait for > % specified peer to initiate a connection with the local system. If > % this does not happen within a specific time (hold time), the local > % system will then also attempt to initiate connection with the > % specified peer. > > > 7) Event 4,5 > > Won't a variant of this with bgp_stop_flap option set be requried ? > We can also achieve the same by using the bgp_stop-Flap option as a > flag that is provided as an input to the state machine. > > > 8) Event 5 Definition > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > Is it correct that this event will occur only when we want to restart > a connection (after it had been terminated due to some reason beside > administrative action) that we had accepted from an > unconfigured peer ? > > > 9) Timer Events Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > > > Event9: connect retry timer expires > > > > Definition: An event triggered by the expiration of=20 > > Event10: hold time expires > > > > Definition: An event generated when the HoldTimer=20 > > We can use any one of these terms (trigerred by/generated > when) for all > timer event definitions to make it look more consistent. > > > > 10) Event 8 Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > Hold Timer is only used when persistent =20 > > BGP oscillation damping functions are=20 > > enabled. =20 > > > > Status: Optional. Used when persistent > > BGP peer oscillation damping functions > > are enabled.=20 > > > The status repeats the senetnce in the definition. Can we remove the > sentence from the Status, or replace with a different definition. > > Do you think the defenition below would fit ? > > % An event trigerred by the expiry of the Idle Hold timer, indicating > % that the session has completed waiting for a back-off period to > % prevent flap, and may now proceed to take any further systemm > % initiated action for this session. > > > -- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id NAA22545 for <idr-archive@nic.merit.edu>; Mon, 9 Dec 2002 13:15:55 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 55CF39126C; Mon, 9 Dec 2002 13:15:08 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 903719126E; Mon, 9 Dec 2002 13:15:07 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 73E699126C for <idr@trapdoor.merit.edu>; Mon, 9 Dec 2002 13:15:05 -0500 (EST) Received: by segue.merit.edu (Postfix) id 294575E061; Mon, 9 Dec 2002 13:15:05 -0500 (EST) Delivered-To: idr@merit.edu Received: from mailgate.pit.comms.marconi.com (mailgate.pit.comms.marconi.com [169.144.68.6]) by segue.merit.edu (Postfix) with ESMTP id 5CFA35DE42 for <idr@merit.edu>; Mon, 9 Dec 2002 13:15:04 -0500 (EST) Received: from mailman.pit.comms.marconi.com (mailman.pit.comms.marconi.com [169.144.2.12]) by mailgate.pit.comms.marconi.com (8.9.3/8.9.3) with ESMTP id NAA08530; Mon, 9 Dec 2002 13:15:01 -0500 (EST) Received: from whq-msgrtr-01.pit.comms.marconi.com (whq-msgrtr-01.pit.comms.marconi.com [169.144.2.221]) by mailman.pit.comms.marconi.com (8.9.3/8.9.3) with ESMTP id NAA20043; Mon, 9 Dec 2002 13:15:02 -0500 (EST) Received: by whq-msgrtr-01.pit.comms.marconi.com with Internet Mail Service (5.5.2650.21) id <X91YVFJA>; Mon, 9 Dec 2002 13:15:01 -0500 Message-ID: <313680C9A886D511A06000204840E1CF3EA336@whq-msgusr-02.pit.comms.marconi.com> From: "Krishnan, Vijay G." <Vijay.G.Krishnan@marconi.com> To: "'PamSri'" <pamsri01@yahoo.com> Cc: "'idr@merit.edu'" <idr@merit.edu> Subject: RE: BGP routes with label information Date: Mon, 9 Dec 2002 13:14:52 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-idr@merit.edu Precedence: bulk Sri, I am sending my reply to IDR as well. Should the label information be considered part of NLRI or an attribute. According to section 3 of RFC3107, label is part of an attribute. "A BGP speaker can withdraw a previously advertised route (as well as the binding between this route and a label) by either (a) advertising a new route (and a label) with the same NLRI as the previously advertised route, or (b) listing the NLRI of the previously advertised route in the Withdrawn Routes field of an Update message. The label information carried (as part of NLRI) in the Withdrawn Routes field should be set to 0x800000." In section 4, it says that we can withdraw a particular label (like an NLRI) "In the case where a BGP speaker advertises multiple routes to a destination, if a route is withdrawn, and a label(s) is specified at the time of withdrawal, only the corresponding route with the corresponding label is withdrawn." So if we need to change the label mapping should we send the new set of labels or withdraw it before sending the new update. thanks Vijay -----Original Message----- From: PamSri [mailto:pamsri01@yahoo.com] Sent: Monday, December 09, 2002 11:48 AM To: Krishnan, Vijay G. Subject: Re: BGP routes with label information Vijay , Can you provide some more info on scenario which causes 2 updates to be sent out back with different label info ? Normally if you encounter a situation whereby you send 2 different sets of path attribs separately then you overwrite the path attribs with the latest update. sri --- "Krishnan, Vijay G." <Vijay.G.Krishnan@marconi.com> wrote: > Hi, > > I have a question regarding multiple label > advertisements for BGP route > (RFC-3107). > > Suppose a BGP speaker receives in a single update > message, a destination > prefix (p) with a nexthop (n) and multiple mpls > labels (L1, L2 and L3). In > the next update, it receives prefix (p) and nexthop > (n) but a different > label (L4). > > Will (p,n,L4) replace {(p,n,L1), (p,n,L2) and > (p,n,L3)}. > Or > Will the new one be added to get {(p,n,L1), > (p,n,L2), (p,n,L3) and > (p,n,L4)}. > > thanks > Vijay > > __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id LAA17220 for <idr-archive@nic.merit.edu>; Mon, 9 Dec 2002 11:05:32 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id AF8439127A; Mon, 9 Dec 2002 11:04:23 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 76AC891269; Mon, 9 Dec 2002 11:04:23 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 9BE349127A for <idr@trapdoor.merit.edu>; Mon, 9 Dec 2002 11:02:13 -0500 (EST) Received: by segue.merit.edu (Postfix) id 864395E056; Mon, 9 Dec 2002 11:02:13 -0500 (EST) Delivered-To: idr@merit.edu Received: from mailgate.pit.comms.marconi.com (mailgate.pit.comms.marconi.com [169.144.68.6]) by segue.merit.edu (Postfix) with ESMTP id 369E35DDC9 for <idr@merit.edu>; Mon, 9 Dec 2002 11:02:13 -0500 (EST) Received: from mailman.pit.comms.marconi.com (mailman.pit.comms.marconi.com [169.144.2.12]) by mailgate.pit.comms.marconi.com (8.9.3/8.9.3) with ESMTP id LAA26346; Mon, 9 Dec 2002 11:01:47 -0500 (EST) Received: from whq-msgrtr-01.pit.comms.marconi.com (whq-msgrtr-01.pit.comms.marconi.com [169.144.2.221]) by mailman.pit.comms.marconi.com (8.9.3/8.9.3) with ESMTP id LAA15771; Mon, 9 Dec 2002 11:01:48 -0500 (EST) Received: by whq-msgrtr-01.pit.comms.marconi.com with Internet Mail Service (5.5.2650.21) id <X91Y4742>; Mon, 9 Dec 2002 11:01:47 -0500 Message-ID: <313680C9A886D511A06000204840E1CF3EA335@whq-msgusr-02.pit.comms.marconi.com> From: "Krishnan, Vijay G." <Vijay.G.Krishnan@marconi.com> To: "'idr@merit.edu'" <idr@merit.edu> Cc: "'yakov@juniper.net'" <yakov@juniper.net>, "'erosen@cisco.com'" <erosen@cisco.com> Subject: BGP routes with label information Date: Mon, 9 Dec 2002 11:01:45 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-idr@merit.edu Precedence: bulk Hi, I have a question regarding multiple label advertisements for BGP route (RFC-3107). Suppose a BGP speaker receives in a single update message, a destination prefix (p) with a nexthop (n) and multiple mpls labels (L1, L2 and L3). In the next update, it receives prefix (p) and nexthop (n) but a different label (L4). Will (p,n,L4) replace {(p,n,L1), (p,n,L2) and (p,n,L3)}. Or Will the new one be added to get {(p,n,L1), (p,n,L2), (p,n,L3) and (p,n,L4)}. thanks Vijay Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id IAA08976 for <idr-archive@nic.merit.edu>; Mon, 9 Dec 2002 08:05:22 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 0E5F09123A; Mon, 9 Dec 2002 08:04:54 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id C838F91259; Mon, 9 Dec 2002 08:04:53 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 790009123A for <idr@trapdoor.merit.edu>; Mon, 9 Dec 2002 08:04:52 -0500 (EST) Received: by segue.merit.edu (Postfix) id 559A45DDC9; Mon, 9 Dec 2002 08:04:52 -0500 (EST) Delivered-To: idr@merit.edu Received: from ietf.org (odin.ietf.org [132.151.1.176]) by segue.merit.edu (Postfix) with ESMTP id CC0EC5DDA8 for <idr@merit.edu>; Mon, 9 Dec 2002 08:04:51 -0500 (EST) Received: from CNRI.Reston.VA.US (localhost [127.0.0.1]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA07024; Mon, 9 Dec 2002 08:01:56 -0500 (EST) Message-Id: <200212091301.IAA07024@ietf.org> Mime-Version: 1.0 Content-Type: Multipart/Mixed; Boundary="NextPart" To: IETF-Announce: ; Cc: idr@merit.edu From: Internet-Drafts@ietf.org Reply-To: Internet-Drafts@ietf.org Subject: I-D ACTION:draft-ietf-idr-bgp-gr-survey-00.txt Date: Mon, 09 Dec 2002 08:01:56 -0500 Sender: owner-idr@merit.edu Precedence: bulk --NextPart A New Internet-Draft is available from the on-line Internet-Drafts directories. This draft is a work item of the Inter-Domain Routing Working Group of the IETF. Title : BGP Graceful Restart - Implementation Survey Author(s) : J. Scudder Filename : draft-ietf-idr-bgp-gr-survey-00.txt Pages : 10 Date : 2002-12-6 This document provides a survey of BGP-4 Graceful Restart implementations. A URL for this Internet-Draft is: http://www.ietf.org/internet-drafts/draft-ietf-idr-bgp-gr-survey-00.txt To remove yourself from the IETF Announcement list, send a message to ietf-announce-request with the word unsubscribe in the body of the message. Internet-Drafts are also available by anonymous FTP. Login with the username "anonymous" and a password of your e-mail address. After logging in, type "cd internet-drafts" and then "get draft-ietf-idr-bgp-gr-survey-00.txt". A list of Internet-Drafts directories can be found in http://www.ietf.org/shadow.html or ftp://ftp.ietf.org/ietf/1shadow-sites.txt Internet-Drafts can also be obtained by e-mail. Send a message to: mailserv@ietf.org. In the body type: "FILE /internet-drafts/draft-ietf-idr-bgp-gr-survey-00.txt". NOTE: The mail server at ietf.org can return the document in MIME-encoded form by using the "mpack" utility. To use this feature, insert the command "ENCODING mime" before the "FILE" command. To decode the response(s), you will need "munpack" or a MIME-compliant mail reader. Different MIME-compliant mail readers exhibit different behavior, especially when dealing with "multipart" MIME messages (i.e. documents which have been split up into multiple messages), so check your local documentation on how to manipulate these messages. Below is the data which will enable a MIME compliant mail reader implementation to automatically retrieve the ASCII version of the Internet-Draft. --NextPart Content-Type: Multipart/Alternative; Boundary="OtherAccess" --OtherAccess Content-Type: Message/External-body; access-type="mail-server"; server="mailserv@ietf.org" Content-Type: text/plain Content-ID: <2002-12-6140149.I-D@ietf.org> ENCODING mime FILE /internet-drafts/draft-ietf-idr-bgp-gr-survey-00.txt --OtherAccess Content-Type: Message/External-body; name="draft-ietf-idr-bgp-gr-survey-00.txt"; site="ftp.ietf.org"; access-type="anon-ftp"; directory="internet-drafts" Content-Type: text/plain Content-ID: <2002-12-6140149.I-D@ietf.org> --OtherAccess-- --NextPart-- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id QAA08701 for <idr-archive@nic.merit.edu>; Fri, 6 Dec 2002 16:05:04 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 0566191306; Fri, 6 Dec 2002 16:03:40 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id C8BD19130B; Fri, 6 Dec 2002 16:03:39 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id EE97691306 for <idr@trapdoor.merit.edu>; Fri, 6 Dec 2002 16:03:33 -0500 (EST) Received: by segue.merit.edu (Postfix) id DC9335DEE1; Fri, 6 Dec 2002 16:03:33 -0500 (EST) Delivered-To: idr@merit.edu Received: from shockwave.systems.pipex.net (shockwave.systems.pipex.net [62.241.160.9]) by segue.merit.edu (Postfix) with ESMTP id AB4EA5DE84 for <idr@merit.edu>; Fri, 6 Dec 2002 16:03:33 -0500 (EST) Received: from tom3 (usercq28.uk.uudial.com [62.188.156.156]) by shockwave.systems.pipex.net (Postfix) with SMTP id 5B9901600102C; Fri, 6 Dec 2002 21:03:30 +0000 (GMT) Message-ID: <004a01c29d6a$e5acda80$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: "Susan Hares" <shares@nexthop.com>, "Natale, Jonathan" <JNatale@celoxnetworks.com>, "idr" <idr@merit.edu> Subject: Re: damping persistent peer oscillations Date: Fri, 6 Dec 2002 21:02:22 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MIMEOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk ** comment inline -----Original Message----- From: Susan Hares <shares@nexthop.com> To: Natale, Jonathan <JNatale@celoxnetworks.com>; idr@merit.edu <idr@merit.edu> Date: 06 December 2002 19:18 Subject: RE: damping persistent peer oscillations Jonathan: The draft is below. I'll be updating the draft-hares-bgp-backoff-00.txt draft this month. Please send comments. The base specification will simply state "damping persistent BGP adjacency flapping ** aka peer oscillation damping :-) are outside the scope of this document. Sue --------------------- Internet Draft S. Hares Document: draft-hares-bgp-backoff-00.txt NextHop Technologies, Inc. Expires: December 2002 June 2002 BGP Peer Restart Backoff Mechanisms -----Original Message----- From: Natale, Jonathan [mailto:JNatale@celoxnetworks.com] Sent: Friday, December 06, 2002 10:47 AM To: idr@merit.edu Subject: damping persistent peer oscillations >From Draft 18: "Local system automatically starts the BGP peer connection with persistent peer oscillation damping enabled. The exact method of damping persistent peer oscillations is left up to the implementation. These methods of damping persistent BGP adjacency flapping are outside the scope of this document." -- is there a draft describing a "method of damping persistent peer oscillations" ? Tom Petch nwnetworks@dial.pipex.com Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id OAA02729 for <idr-archive@nic.merit.edu>; Fri, 6 Dec 2002 14:18:55 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id DC4E0912EC; Fri, 6 Dec 2002 14:18:43 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id ADC88912ED; Fri, 6 Dec 2002 14:18:43 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 215DE912EC for <idr@trapdoor.merit.edu>; Fri, 6 Dec 2002 14:17:51 -0500 (EST) Received: by segue.merit.edu (Postfix) id F13345DE95; Fri, 6 Dec 2002 14:17:50 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 945DD5DE8B for <idr@merit.edu>; Fri, 6 Dec 2002 14:17:50 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB6JHnG92226 for idr@merit.edu; Fri, 6 Dec 2002 14:17:49 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB6JHhC92212 for <idr@merit.edu>; Fri, 6 Dec 2002 14:17:43 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Subject: RE: damping persistent peer oscillations X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 Date: Fri, 6 Dec 2002 14:17:42 -0500 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA61690@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: damping persistent peer oscillations Thread-Index: AcKdPu5Bw+ERPh5gSV6yMqSEWwKbfAAHPs+w From: "Susan Hares" <shares@nexthop.com> To: "Natale, Jonathan" <JNatale@celoxnetworks.com>, <idr@merit.edu> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id OAA02729 Jonathan: The draft is below. I'll be updating the draft-hares-bgp-backoff-00.txt draft this month. Please send comments. The base specification will simply state "damping persistent BGP adjacency flapping are outside the scope of this document. Sue --------------------- Internet Draft S. Hares Document: draft-hares-bgp-backoff-00.txt NextHop Technologies, Inc. Expires: December 2002 June 2002 BGP Peer Restart Backoff Mechanisms -----Original Message----- From: Natale, Jonathan [mailto:JNatale@celoxnetworks.com] Sent: Friday, December 06, 2002 10:47 AM To: idr@merit.edu Subject: damping persistent peer oscillations >From Draft 18: "Local system automatically starts the BGP peer connection with persistent peer oscillation damping enabled. The exact method of damping persistent peer oscillations is left up to the implementation. These methods of damping persistent BGP adjacency flapping are outside the scope of this document." -- is there a draft describing a "method of damping persistent peer oscillations" ? Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id OAA02492 for <idr-archive@nic.merit.edu>; Fri, 6 Dec 2002 14:14:21 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 28237912EB; Fri, 6 Dec 2002 14:13:45 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id E37C4912EC; Fri, 6 Dec 2002 14:13:44 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 7E7CF912EB for <idr@trapdoor.merit.edu>; Fri, 6 Dec 2002 14:13:43 -0500 (EST) Received: by segue.merit.edu (Postfix) id 6D0CE5DFB7; Fri, 6 Dec 2002 14:13:43 -0500 (EST) Delivered-To: idr@merit.edu Received: from workhorse.fictitious.org (workhorse.fictitious.org [209.150.1.230]) by segue.merit.edu (Postfix) with ESMTP id 3AB545DFBB for <idr@merit.edu>; Fri, 6 Dec 2002 14:13:42 -0500 (EST) Received: from workhorse.fictitious.org (localhost.fictitious.org [127.0.0.1]) by workhorse.fictitious.org (8.9.3/8.9.3) with ESMTP id OAA22088; Fri, 6 Dec 2002 14:12:31 -0500 (EST) (envelope-from curtis@workhorse.fictitious.org) Message-Id: <200212061912.OAA22088@workhorse.fictitious.org> To: "Natale, Jonathan" <JNatale@celoxnetworks.com> Cc: "'curtis@fictitious.org'" <curtis@fictitious.org>, "'idr@merit.edu'" <idr@merit.edu> Reply-To: curtis@fictitious.org Subject: Re: damping persistent peer oscillations In-reply-to: Your message of "Fri, 06 Dec 2002 14:01:10 EST." <1117F7D44159934FB116E36F4ABF221B02C7C6DF@celox-ma1-ems1.celoxnetworks.com> Date: Fri, 06 Dec 2002 14:12:31 -0500 From: Curtis Villamizar <curtis@fictitious.org> Sender: owner-idr@merit.edu Precedence: bulk In message <1117F7D44159934FB116E36F4ABF221B02C7C6DF@celox-ma1-ems1.celoxnetwor ks.com>, "Natale, Jonathan" writes: > Curtis, > > Thanks for the reply. > > But I was looking for *peer* oscillation damping, not *route* oscillation > damping. RFC1771 had: > > "For a peer that was previously transitioned to Idle due to an error, > the time between consecutive generation of Start events, if such events are > generated automatically, shall exponentially increase." The text I quoted started with When a peer routing session is broken, ... and suggested the peering session itself may be marked as unstable. How is this not applicable? > --Current implementations (AKA Juniper and Cisco) do not do this, and the > draft-18 has it tagged as out of scope (see below). > > Maybe this will be in the RFC1772 re-do and/or a BCP? Since its out of scope but OK to do something as stated in -18, its OK to do what is recommended in RFC2439 but not required. You asked if anything addressed a "method of damping persistent peer oscillations". The answer is "yes there is". Curtis > > -----Original Message----- > > From: Curtis Villamizar [mailto:curtis@fictitious.org] > > Sent: Friday, December 06, 2002 1:33 PM > > To: Natale, Jonathan > > Cc: idr@merit.edu > > Subject: Re: damping persistent peer oscillations > > > > > > > > In message > > <1117F7D44159934FB116E36F4ABF221B02C7C6D7@celox-ma1-ems1.celoxnetwor > > ks.com>, "Natale, Jonathan" writes: > > > >From Draft 18: > > > "Local system automatically starts the BGP peer connection > > with persistent > > > peer oscillation damping enabled. The exact method of > > damping persistent > > > peer oscillations is left up to the implementation. These > > methods of > > > damping persistent BGP adjacency flapping are outside the > > scope of this > > > document." > > > > > > -- is there a draft describing a "method of damping persistent peer > > > oscillations" ? > > > > > > RFC 2439 "BGP Route Flap Damping": > > > > 4.8.5 Processing A Peer Router Loss > > > > When a peer routing session is broken, either all individual routes > > advertised by that peer may be marked as unstable, or the peering > > session itself may be marked as unstable. Marking the > > peer will save > > considerable memory. Since the individual routes are advertised as > > unreachable to routers beyond the immediate problem, per > > route state > > will be incurred beyond the peer immediately adjacent to the BGP > > session that went down. If the instability continues, the > > immediately adjacent router need only keep track of the peer > > stability history. The routers beyond that point will receive no > > further advertisements or withdrawal of routes and will dispose of > > the damping structure over time. > > > > BGP notification through an optional transitive attribute that > > damping will already be applied may be considered in the future to > > reduce the number of routers that incur damping structure storage > > > > Curtis > > > Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id OAA02318 for <idr-archive@nic.merit.edu>; Fri, 6 Dec 2002 14:11:13 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id BC241912EA; Fri, 6 Dec 2002 14:10:40 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 8357A912EB; Fri, 6 Dec 2002 14:10:40 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 08041912EA for <idr@trapdoor.merit.edu>; Fri, 6 Dec 2002 14:10:39 -0500 (EST) Received: by segue.merit.edu (Postfix) id E80655DDC8; Fri, 6 Dec 2002 14:10:38 -0500 (EST) Delivered-To: idr@merit.edu Received: from nemesis.systems.pipex.net (nemesis.systems.pipex.net [62.241.160.8]) by segue.merit.edu (Postfix) with ESMTP id 9A8115DFB6 for <idr@merit.edu>; Fri, 6 Dec 2002 14:10:38 -0500 (EST) Received: from tom3 (userbm41.uk.uudial.com [62.188.144.247]) by nemesis.systems.pipex.net (Postfix) with SMTP id 9C34416008098; Fri, 6 Dec 2002 19:10:32 +0000 (GMT) Message-ID: <00e701c29d5b$1f5bf460$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: <curtis@fictitious.org>, "Natale, Jonathan" <JNatale@celoxnetworks.com> Cc: "idr" <idr@merit.edu> Subject: Re: damping persistent peer oscillations Date: Fri, 6 Dec 2002 19:10:15 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MIMEOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Whoops; I asked for the word flapping to be taken out of this section because it had nothing to do with this RFC and might confuse people into thinking it had! (And my comment was accepted by Yakov). My understanding is different, that peer oscillation damping refers to IDs such as draft-hares-bgp-backoff-01.txt which appeared in the summer and which are on hold, to be returned to once the base spec has been finalised. Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Curtis Villamizar <curtis@fictitious.org> To: Natale, Jonathan <JNatale@celoxnetworks.com> Cc: idr@merit.edu <idr@merit.edu> Date: 06 December 2002 18:34 Subject: Re: damping persistent peer oscillations > >In message <1117F7D44159934FB116E36F4ABF221B02C7C6D7@celox-ma1-ems1.celoxnetwor >ks.com>, "Natale, Jonathan" writes: >> >From Draft 18: >> "Local system automatically starts the BGP peer connection with persistent >> peer oscillation damping enabled. The exact method of damping persistent >> peer oscillations is left up to the implementation. These methods of >> damping persistent BGP adjacency flapping are outside the scope of this >> document." >> >> -- is there a draft describing a "method of damping persistent peer >> oscillations" ? > > >RFC 2439 "BGP Route Flap Damping": > > 4.8.5 Processing A Peer Router Loss > > When a peer routing session is broken, either all individual routes > advertised by that peer may be marked as unstable, or the peering > session itself may be marked as unstable. Marking the peer will save > considerable memory. Since the individual routes are advertised as > unreachable to routers beyond the immediate problem, per route state > will be incurred beyond the peer immediately adjacent to the BGP > session that went down. If the instability continues, the > immediately adjacent router need only keep track of the peer > stability history. The routers beyond that point will receive no > further advertisements or withdrawal of routes and will dispose of > the damping structure over time. > > BGP notification through an optional transitive attribute that > damping will already be applied may be considered in the future to > reduce the number of routers that incur damping structure storage > >Curtis Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id OAA01778 for <idr-archive@nic.merit.edu>; Fri, 6 Dec 2002 14:02:02 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id CBC06912E7; Fri, 6 Dec 2002 14:01:34 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 81A7F912E8; Fri, 6 Dec 2002 14:01:32 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 79BA6912E5 for <idr@trapdoor.merit.edu>; Fri, 6 Dec 2002 14:01:26 -0500 (EST) Received: by segue.merit.edu (Postfix) id 88DDD5DF61; Fri, 6 Dec 2002 14:01:11 -0500 (EST) Delivered-To: idr@merit.edu Received: from celox-ma1-ems1.celoxnetworks.com (celox-ma1-imap1.celoxnetworks.com [12.40.60.233]) by segue.merit.edu (Postfix) with ESMTP id 4C0A55DDC8 for <idr@merit.edu>; Fri, 6 Dec 2002 14:01:11 -0500 (EST) Received: by celox-ma1-ems1.celoxnetworks.com with Internet Mail Service (5.5.2653.19) id <YL0CTXQY>; Fri, 6 Dec 2002 14:01:11 -0500 Message-ID: <1117F7D44159934FB116E36F4ABF221B02C7C6DF@celox-ma1-ems1.celoxnetworks.com> From: "Natale, Jonathan" <JNatale@celoxnetworks.com> To: "'curtis@fictitious.org'" <curtis@fictitious.org> Cc: "'idr@merit.edu'" <idr@merit.edu> Subject: RE: damping persistent peer oscillations Date: Fri, 6 Dec 2002 14:01:10 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain Sender: owner-idr@merit.edu Precedence: bulk Curtis, Thanks for the reply. But I was looking for *peer* oscillation damping, not *route* oscillation damping. RFC1771 had: "For a peer that was previously transitioned to Idle due to an error, the time between consecutive generation of Start events, if such events are generated automatically, shall exponentially increase." --Current implementations (AKA Juniper and Cisco) do not do this, and the draft-18 has it tagged as out of scope (see below). Maybe this will be in the RFC1772 re-do and/or a BCP? > -----Original Message----- > From: Curtis Villamizar [mailto:curtis@fictitious.org] > Sent: Friday, December 06, 2002 1:33 PM > To: Natale, Jonathan > Cc: idr@merit.edu > Subject: Re: damping persistent peer oscillations > > > > In message > <1117F7D44159934FB116E36F4ABF221B02C7C6D7@celox-ma1-ems1.celoxnetwor > ks.com>, "Natale, Jonathan" writes: > > >From Draft 18: > > "Local system automatically starts the BGP peer connection > with persistent > > peer oscillation damping enabled. The exact method of > damping persistent > > peer oscillations is left up to the implementation. These > methods of > > damping persistent BGP adjacency flapping are outside the > scope of this > > document." > > > > -- is there a draft describing a "method of damping persistent peer > > oscillations" ? > > > RFC 2439 "BGP Route Flap Damping": > > 4.8.5 Processing A Peer Router Loss > > When a peer routing session is broken, either all individual routes > advertised by that peer may be marked as unstable, or the peering > session itself may be marked as unstable. Marking the > peer will save > considerable memory. Since the individual routes are advertised as > unreachable to routers beyond the immediate problem, per > route state > will be incurred beyond the peer immediately adjacent to the BGP > session that went down. If the instability continues, the > immediately adjacent router need only keep track of the peer > stability history. The routers beyond that point will receive no > further advertisements or withdrawal of routes and will dispose of > the damping structure over time. > > BGP notification through an optional transitive attribute that > damping will already be applied may be considered in the future to > reduce the number of routers that incur damping structure storage > > Curtis > Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id NAA00246 for <idr-archive@nic.merit.edu>; Fri, 6 Dec 2002 13:34:28 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 2D694912E3; Fri, 6 Dec 2002 13:34:14 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id EEE43912E4; Fri, 6 Dec 2002 13:34:13 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id AA5A9912E3 for <idr@trapdoor.merit.edu>; Fri, 6 Dec 2002 13:34:12 -0500 (EST) Received: by segue.merit.edu (Postfix) id 882905DF61; Fri, 6 Dec 2002 13:34:12 -0500 (EST) Delivered-To: idr@merit.edu Received: from workhorse.fictitious.org (workhorse.fictitious.org [209.150.1.230]) by segue.merit.edu (Postfix) with ESMTP id C8BB75DE6B for <idr@merit.edu>; Fri, 6 Dec 2002 13:34:11 -0500 (EST) Received: from workhorse.fictitious.org (localhost.fictitious.org [127.0.0.1]) by workhorse.fictitious.org (8.9.3/8.9.3) with ESMTP id NAA21255; Fri, 6 Dec 2002 13:33:01 -0500 (EST) (envelope-from curtis@workhorse.fictitious.org) Message-Id: <200212061833.NAA21255@workhorse.fictitious.org> To: "Natale, Jonathan" <JNatale@celoxnetworks.com> Cc: idr@merit.edu Reply-To: curtis@fictitious.org Subject: Re: damping persistent peer oscillations In-reply-to: Your message of "Fri, 06 Dec 2002 10:47:19 EST." <1117F7D44159934FB116E36F4ABF221B02C7C6D7@celox-ma1-ems1.celoxnetworks.com> Date: Fri, 06 Dec 2002 13:33:01 -0500 From: Curtis Villamizar <curtis@fictitious.org> Sender: owner-idr@merit.edu Precedence: bulk In message <1117F7D44159934FB116E36F4ABF221B02C7C6D7@celox-ma1-ems1.celoxnetwor ks.com>, "Natale, Jonathan" writes: > >From Draft 18: > "Local system automatically starts the BGP peer connection with persistent > peer oscillation damping enabled. The exact method of damping persistent > peer oscillations is left up to the implementation. These methods of > damping persistent BGP adjacency flapping are outside the scope of this > document." > > -- is there a draft describing a "method of damping persistent peer > oscillations" ? RFC 2439 "BGP Route Flap Damping": 4.8.5 Processing A Peer Router Loss When a peer routing session is broken, either all individual routes advertised by that peer may be marked as unstable, or the peering session itself may be marked as unstable. Marking the peer will save considerable memory. Since the individual routes are advertised as unreachable to routers beyond the immediate problem, per route state will be incurred beyond the peer immediately adjacent to the BGP session that went down. If the instability continues, the immediately adjacent router need only keep track of the peer stability history. The routers beyond that point will receive no further advertisements or withdrawal of routes and will dispose of the damping structure over time. BGP notification through an optional transitive attribute that damping will already be applied may be considered in the future to reduce the number of routers that incur damping structure storage Curtis Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id KAA21022 for <idr-archive@nic.merit.edu>; Fri, 6 Dec 2002 10:48:39 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id BEE889124F; Fri, 6 Dec 2002 10:48:18 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 92B5491291; Fri, 6 Dec 2002 10:48:18 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 015179124F for <idr@trapdoor.merit.edu>; Fri, 6 Dec 2002 10:47:28 -0500 (EST) Received: by segue.merit.edu (Postfix) id BF2B45DF1A; Fri, 6 Dec 2002 10:47:28 -0500 (EST) Delivered-To: idr@merit.edu Received: from celox-ma1-ems1.celoxnetworks.com (celox-ma1-imap1.celoxnetworks.com [12.40.60.233]) by segue.merit.edu (Postfix) with ESMTP id 194105DF20 for <idr@merit.edu>; Fri, 6 Dec 2002 10:47:21 -0500 (EST) Received: by celox-ma1-ems1.celoxnetworks.com with Internet Mail Service (5.5.2653.19) id <YL0CTWRR>; Fri, 6 Dec 2002 10:47:20 -0500 Message-ID: <1117F7D44159934FB116E36F4ABF221B02C7C6D7@celox-ma1-ems1.celoxnetworks.com> From: "Natale, Jonathan" <JNatale@celoxnetworks.com> To: idr@merit.edu Subject: damping persistent peer oscillations Date: Fri, 6 Dec 2002 10:47:19 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain Sender: owner-idr@merit.edu Precedence: bulk >From Draft 18: "Local system automatically starts the BGP peer connection with persistent peer oscillation damping enabled. The exact method of damping persistent peer oscillations is left up to the implementation. These methods of damping persistent BGP adjacency flapping are outside the scope of this document." -- is there a draft describing a "method of damping persistent peer oscillations" ? Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id GAA05552 for <idr-archive@nic.merit.edu>; Fri, 6 Dec 2002 06:05:33 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 480F79121A; Fri, 6 Dec 2002 06:03:29 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 019649123F; Fri, 6 Dec 2002 06:03:28 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 0FE849121A for <idr@trapdoor.merit.edu>; Fri, 6 Dec 2002 06:03:05 -0500 (EST) Received: by segue.merit.edu (Postfix) id E25815DE6C; Fri, 6 Dec 2002 06:03:05 -0500 (EST) Delivered-To: idr@merit.edu Received: from ganesh.ctd.hctech.com (unknown [202.54.64.2]) by segue.merit.edu (Postfix) with ESMTP id 2A1565DD9B for <idr@merit.edu>; Fri, 6 Dec 2002 06:03:02 -0500 (EST) Received: by GANESH with Internet Mail Service (5.5.2653.19) id <YFVC170G>; Fri, 6 Dec 2002 16:41:17 +0530 Message-ID: <55E277B99171E041ABF5F4B1C6DDCA06E4ECAA@haritha.hclt.com> From: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com> To: Susan Hares <shares@nexthop.com>, Yakov Rekhter <yakov@juniper.net>, Jeffrey Haas <jhaas@nexthop.com> Cc: idr@merit.edu, "Sivananda Ramnath (E-mail)" <siva@ctd.hcltech.com>, "Venu Kumar G - CTD, Chennai." <venug@ctd.hcltech.com> Subject: RE: Response to FSM input - Comments 1-10 Date: Fri, 6 Dec 2002 16:35:16 +0530 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-idr@merit.edu Precedence: bulk Hello, Revised summary: Comment 1 - No text change, Potential consensus Comment 2a - No consensus Comment 2b - No consensus Comment 3 - New text suggested, no consensus Comment 4 - Resolved, text changed. Can we consider this closed ? Comment 5 - No text change, Potential consensus Comment 6 - No consensus Comment 7 - No consensus, comment re-phrased. Comment 8 - No text change, potential consensus Comment 9 - Resolved, text changed. Can we consider this closed ? Comment 10 - Resolved, text changed. Can we consider this closed ? > > Comment 1 - no consensus > Comment 2a - out of date (No IdleHold Timer exists) > Comment 2b - no consensus, can't remove due to OpenSent state > action. > > Comment 3 - no text proposed, do we want all options > specified in 8.1 (7 options) > Please review and comment on text > > Comment 4 - agree > Comment 5 - no text proposed, no consensus > Comment 6 - no consensus on the text > Comment 7 - see comment 3, no consensus > Comment 8 - question, no text > Comment 9 - agree, I supplied text > Comment 10 - agree to remove text in option, > different text for definition. > > > Sue Hares > > ========== > > comment 1: > > 1) removal of peer oscillation damping > > The reason we should not remove the peer oscillation damping > from the state machine: > > 1) Deployed implementations support peer oscillation damping > 2) Hooks for the additions in the FSM cannot be added later. > > These hooks are optional and do not need to be implemented. I understand. I am not trying to object to peer oscillation damping, I think it is a good idea and we have included it in our implementation as well. I was suggesting that instead of a partial description in this draft, it be completely described in the draft on peer oscillation damping. However, I do see your point, and unless there are any objections from others, I think we have consensus on this issue. > > > 2) Idle HoldTimer in draft-18 > > Perhaps this is an old comment. I don't find IdleHoldTimer > in the messages. It is listed as Event 8. Perhaps this is what needs to be removed ? > > 3) HoldTime and HoldTimer > > The Hold Time is specified on page 13 of the draft-18 specification > as to the value and size. The text in OpenSent requires > both a HoldTime and HoldTimer value. > =========== > OpenSent: > > In this state BGP waits for an Open Message from its peer. [SNIP] > > =================== I understand. However, on this basis, we would also need a Keepalive time (since the basis on wihch this is be arrived at can be configurable). I don't know whether we do need to make such changes to the draft, and I am OK to leaving this portion as it stands. > The earlier comments on the list indicated we need to collect what > timers or Timer values were utilized in the FSM. I believe Alex > made this comment last January. (the list archives will have > this information.) > > So, I don't see any way to The DelayBgpOpenTimer is missing from the session attributes list, how about including it ? > > Comment 3: optional parameters specified > > Please suggest text. We have the following optional parameters: > > 1) Delay Open > 2) Accept unconfigured peers > 3) automatic start [Event 3] > 4) passive TCP establishment (optional) > 4 + 5 = Event 5 > 5) BGP_stop_flap (BGP Peer oscillation) [Event 6] > 6)automatic stop [Event 7] > 7)Collision detect in Established State > > If you want some of the optional parameters specified, let's > catch all. Please suggest text and location. It could > go in 8.1. My view is this can be split into two categories: Optional attributes exist per session, and optional attributes for the entire system. How about adding the following to section 8 (immediately before 8.1) % Optional session attributes that a implementation may support % for a session are: % % 1) DelayOpen flag % 2) DelayOpen Timer % 3) Perform automatic start % 4) Passive TCP establishment % 5) Perform automatic stop % 6) BGP Perform % % % Implementations may also support the following optinal attributes % that modify the behavior of the state machine % % 1) Accept un-configured peers % 2) Perform collision detect in Established state I guess it is possible to have some of these things listed under both categories. Am I opening up a can of worms here ? :-) > -------------------------------- > Comment 4 > > Old Text: > Event 1: Manual Start > Definition: Administrator manually starts peer connection. > > ... > Event 2: Manual Stop > Definition: Local system administrator manually stops the > peer connection. > New Text: > Event1: Manual start > Definition: Local system administrator manually starts the > bgp connection. > > ... > Event 2: Manual stop: > Definition: Local system administrator manually stops the > bgp connection. > > agree? Agreed. > --------------- > Comment 5: > > 5) Event3, Event5, Event6 and Event7 > > 5a) > ========= > Event 3/6/7 - cisco indicated the automatic stop, see the NANOG > archives for a discussion of the Bad Route > > http://www.merit.edu/mail.archives/nanog/2001-06/msg00805.html > > The "event" sequences are new with this version of the FSM > to identify particular events by number. I am unclear on what this issue is! > 5b > ====== > > Can we give examples of these events, as implemneted by curent > implementations ? This will be of help to new implementaions. > > An example of Auto Stop is already present in the Update event > handling section, and we can move that to this place as well. > > I believe that examples are outside the scope of this specification. > If you have text clarifying the event descriptions, please send it in. Well, the current text in the established state already gives an example for an automatic start event. I was suggesting that such examples be moved to the event description section. I also think adding text to clarify the intent behind optional events, and giving examples would help, especially for new implementors. However, if you feel this is outside the scope of the RFC, we can leave things as is. > > > ======================================================= > > 6) Event 4,5 Definition > > > Definition: Administrator manually start the peer > > Connection, but has the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing the connection. > > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > > >My understanding is this is used to indicate a peer configuration > >where we start the state machine, and wait for an incoming connection > >request from the configured peer. Can we re-phrase this section to > >indicate this. > > >How about something like: > > > >% The passive flag indicates that the state machine will wait for > >% specified peer to initiate a connection with the local system. If > >% this does not happen within a specific time (hold time), the local > >% system will then also attempt to initiate connection with the > >% specified peer. > > The text in 8.2.1.1 indicates the definition of the passive flag. Yes. > > 6a) > ========== > My understanding of your text is that you want to replace in both > sets of text: > > "The passive flag indicates the peer will listen prior to > establishing a connection". > > with: > > "The passive flag indicates that the state machine will wait for the > specified peer to initiate a connection with a local system. > > The problem with this sentence is that in the "unconfigured" case > the phrase "specified" peer is confusing. I think the original text > is clearer. I am ok to leave the text as is, based on text in 8.2.1.1. However, I do not quite understand your point. In the case of an un-configured peer, the local system cannot 'establish a connection', so is the unconfigured peer case still valid in this context ? > 6b) > ========== > If this does not happen within a specific time (hold time), the local > system will then also attempt to initiate (a) connection with the > specified peer. > > My comments: Again, the "specified peer" term is confusing. Also, > the 2nd half of the statement mixes the actions of the state machine > with the events. I believe this muddies the text instead of > clarifying it. Please refer to my comments for 6a. I think a reason for this is the requirement (in 8.2.1) that a FSM be established a-priori for "accepting each incoming TCP connection for which the peer has not beet identified". Is the intent to establish a FSM session for accepting TCP connections from un-configured peers ? If so, I am of the opinion that this is purely an implementation issue, and need not be a requirement in the spec. > > ================= > Comment 7 - > > Event 6 - (not specified, please confirm) > > > Won't a variant of this with bgp_stop_flap option set be requried ? > > We can also achieve the same by using the bgp_stop-Flap option as a > > flag that is provided as an input to the state machine. > > see comment number 3. bgp_stop_flap is an option. > > This state machine description has many events to describe the > mixture of options and actions. I believe it has a better clarity > than testing lots of flags within a state. > > A proof in case, is that many people are making comments about > specific transitions. I tink I should have phrased my question much better. We already have Event 3 - Automatic Start Event 5 - Automatic start with bgp_stop_flap option set To make things consistent, shouldn't we either a) Add 3 new events : 1) Manual start with bgp_stop flap option set 2) Manual start with passive TCP establishment and bgp_stop_flap option set 3) Automatic start with passive TCP establishment and bgp_stop_flap option set or b) Remove Event 6, and rely on a flag to tell us wether peer flap damping is to be performed for the session or not. > > ============ > Comment 8 - > > > 8) Event 5 Definition > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > > Is it correct that this event will occur only when we want > to restart > > a connection (after it had been terminated due to some reason beside > > administrative action) that we had accepted from an > unconfigured peer ? > > The automatic start function is an implementation specific mechanism. > This text does not seek to restrict it in any fashion. Well, for new implementors, such clarifications will help undertsand how to generate and make use of this event. Again, I am OK with leaving text as is. > > ================ > 9) Timer Events Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > > > Event9: connect retry timer expires > > > > Definition: An event triggered by the expiration of=20 > > Event10: hold time expires > > > > Definition: An event generated when the HoldTimer=20 > > We can use any one of these terms (trigerred by/generated > when) for all > timer event definitions to make it look more consistent. > > agree: Will use "generate" > > Event 8: An event generated when the Idle Hold timer expires. > Event 9: An event generated when the ConnectRetry timer expires. > Event 10: An event generated when the Hold timer expires. > Event 11: An event generated when the Keepalive timer expires > Event 12: An event generated when the Delay BGP Open timer expires. > > Please confirm that this change to the definitions is agreeable to > you. Agreed. I think we have consensus on this issue, unless anyone else has any comments on this. > > ================ > 10) Event 8 Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > Hold Timer is only used when persistent =20 > > BGP oscillation damping functions are=20 > > enabled. =20 > > > > Status: Optional. Used when persistent > > BGP peer oscillation damping functions > > are enabled.=20 > > > The status repeats the senetnce in the definition. Can we remove the > sentence from the Status, or replace with a different definition. > > Do you think the defenition below would fit ? > > % An event trigerred by the expiry of the Idle Hold timer, indicating > % that the session has completed waiting for a back-off period to > % prevent flap, and may now proceed to take any further systemm > % initiated action for this session. > > text change suggestion for 1st sentence that I'll agree to. > > An event trigerred by the expiry of the Idle Hold timer, indicating > that the session has completed waiting for a back-off period to > prevent bgp peer oscillation. > > Please confirm that you will agree to this. Agreed. I think we have consensus on this issue, unless anyone else has any comments on this. Siva (siva@ctd.hcltech.com)> Original text of comments 1-10 follow: > > ------- Message 4 > > Date: Fri, 29 Nov 2002 19:46:06 +0530 > From: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com> > To: skh@nexthop.com, Yakov Rekhter <yakov@juniper.net> > cc: skh@ndzh.com > Subject: Comments on draft 18 > > This message is in MIME format. Since your mail reader does > not understand > this format, some or all of this message may not be legible. > > - ------_=_NextPart_000_01C297B1.DB0F1700 > Content-Type: text/plain; > charset="iso-8859-1" > > Hello, > > Attached are some comments that I have on draft-ietf-idr-bgp4-18. > > Please have a look at these and tell me if the same can > be raised to the > mailing list as well. > > Thanks, > Siva > (siva@ctd.hcltech.com) > > > > - ------_=_NextPart_000_01C297B1.DB0F1700 > Content-Type: text/plain; > name="BGP_Issues.txt" > Content-Transfer-Encoding: quoted-printable > Content-Disposition: attachment; > filename="BGP_Issues.txt" > > All quoted sections of the original document have been > prefixed by a '> = > ' > All lines of proposed text have been prefixed by a '% ' > > > 1) Performing of peer oscillation damping: > > Since this feature is going to be added in a new draft, and its > addition will change the operation of the state machine, can we remove > all mention of it in the state machine ? As part of this removal, can > we also remove the IdleHold Timer from the FSM since it is not useful > in the absence of peer oscillation damping ? > > The draft that describes this procedure can then describe the change > in the state machine required to do this. > > > 2) Session Attributes: > > > Session Attributes required for each connection are;=20 > > > 1) state > > 2) Connect Retry timer > > 3) Hold timer > > 4) Hold time > > 5) Keepalive timer > > Why have a Hold Timer and a Hold Time ? Can we replace this with just > Hold timer ? > > Can we also add the following session attributes: > > a) DelayBgpOpenTimer > b) IdleHold Timer (in case we choose not to remove this from the bas = > FSM) > > Can we also add the following flag to the session attributes: > a) DelayOpen Flag > > > 3) All sessions attributes: > > There are some attributes that are common to all sessions of the = > system. > Can we call these out as well: > > * Accept Connections from unconfigured peers (Enabled/Disabled) > * Peer Oscillation Dampening (Enabled/Disabled) (In case we choose > not to remove it from base spec) > > > 4) Description for Event1/Event2 > > > Event1: Manual start=20 > > > > Definition: Administrator manually starts peer > > connection > > > Event2: Manual stop > > > Definition: Local system administrator manually=20 > > stops the peer connection. > > We can make these look consistent by using the Term Administrator or > Local System administrator in both cases. > > > 5) Event3, Event5, Event6 and Event7 > > Can we give examples of these events, as implemneted by curent > implementations ? This will be of help to new implementaions. > > An example of Auto Stop is already present in the Update event > handling section, and we can move that to this place as well. > > > 6) Event 4,5 Definition > > > Definition: Administrator manually start the peer > > Connection, but has the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing the connection. > > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > > My understanding is this is used to indicate a peer configuration > where we start the state machine, and wait for an incoming connection > request from the configured peer. Can we re-phrase this section to > indicate this. > > How about something like: > > % The passive flag indicates that the state machine will wait for > % specified peer to initiate a connection with the local system. If > % this does not happen within a specific time (hold time), the local > % system will then also attempt to initiate connection with the > % specified peer. > > > 7) Event 4,5 > > Won't a variant of this with bgp_stop_flap option set be requried ? > We can also achieve the same by using the bgp_stop-Flap option as a > flag that is provided as an input to the state machine. > > > 8) Event 5 Definition > > > Definition: Local system automatically starts the > > BGP session with the passive flag=20 > > enabled. The passive flag indicates=20 > > that the peer will listen prior to=20 > > establishing a connection. > > Is it correct that this event will occur only when we want to restart > a connection (after it had been terminated due to some reason beside > administrative action) that we had accepted from an > unconfigured peer ? > > > 9) Timer Events Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > > > Event9: connect retry timer expires > > > > Definition: An event triggered by the expiration of=20 > > Event10: hold time expires > > > > Definition: An event generated when the HoldTimer=20 > > We can use any one of these terms (trigerred by/generated > when) for all > timer event definitions to make it look more consistent. > > > > 10) Event 8 Definition > > > Event8: idle Hold timer expires=20 > > > > Definition: Idle Hold timer expires. The Idle=20 > > Hold Timer is only used when persistent =20 > > BGP oscillation damping functions are=20 > > enabled. =20 > > > > Status: Optional. Used when persistent > > BGP peer oscillation damping functions > > are enabled.=20 > > > The status repeats the senetnce in the definition. Can we remove the > sentence from the Status, or replace with a different definition. > > Do you think the defenition below would fit ? > > % An event trigerred by the expiry of the Idle Hold timer, indicating > % that the session has completed waiting for a back-off period to > % prevent flap, and may now proceed to take any further systemm > % initiated action for this session. > > > -- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id TAA01446 for <idr-archive@nic.merit.edu>; Thu, 5 Dec 2002 19:01:49 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 17F59912D8; Thu, 5 Dec 2002 19:00:29 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id DFE2A912D9; Thu, 5 Dec 2002 19:00:28 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 1450F912D8 for <idr@trapdoor.merit.edu>; Thu, 5 Dec 2002 19:00:25 -0500 (EST) Received: by segue.merit.edu (Postfix) id B0F505DF3A; Thu, 5 Dec 2002 19:00:22 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id B51305E0FB for <idr@merit.edu>; Thu, 5 Dec 2002 19:00:20 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB600Jg71542 for idr@merit.edu; Thu, 5 Dec 2002 19:00:19 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB600DC71529 for <idr@merit.edu>; Thu, 5 Dec 2002 19:00:13 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message Subject: RE: bgp18 WG Last Call fsm missing next state MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 Date: Thu, 5 Dec 2002 19:00:13 -0500 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AB10@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: bgp18 WG Last Call fsm missing next state Thread-Index: AcKcrBbJC1arMGN2RcqWX3lbKoQCwgABrflg From: "Susan Hares" <shares@nexthop.com> To: "Tom Petch" <nwnetworks@dial.pipex.com>, "Jeffrey Haas" <jhaas@nexthop.com> Cc: "idr" <idr@merit.edu> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id TAA01446 Tom: Thanks for the input. 1 - done 2 - sent to the list for implementations 3 - done, and question about 1 addition. 4,5 - I'll look for different thread 6 - sent to list for implementations I'm happy with alternative 2 if no one else works on it. 7 - this mechanism allows a single fsm to handle both. 2 fsm and 1 fsm BGP FSM seem to exist. (I queried implementors a few times on this one. So, I just put in this change to provide the flexibility. Collision detect tends to give scrambled brains for most people.. As Dennis Ferguson said 2 years ago, that's the hardest part of the FSM. Soo.. bottom line. I'll look to see if any implementors work on 2, 6 or 7. I'll query a few of the implementors in a questionnaire as well. Thanks for the hard work! Sue 3's question --------- Should I add completes BGP initialization in the following list: If the TCP connection succeeds [Event 15 or Event 16], local system checks the Delay Open flag prior to processing: If the Delay Open flag is set, the local system: - clears the connect retry timer, %- sets the BGP** Open Delay timer to initial value, and %- complets the BGP initialization - stays in the Connect state. If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clears the BGP**Open Delay timer (sets to zero), - completes the BGP initialization, - sends an Open message to its peer, - sets the hold timer to a large value, and - changes the state to Open Sent. -----Original Message----- From: Tom Petch [mailto:nwnetworks@dial.pipex.com] Sent: Thursday, December 05, 2002 5:15 PM To: Susan Hares; Jeffrey Haas Cc: idr Subject: Re: bgp18 WG Last Call fsm missing next state Sue, Jeff Agree with 1 (with a slight change of words put **inline to make the name the Open Delay timer as Yakov accepted - done 2 wait on implementors - but do we need a then clause for when peer oscillation damping does not allow a new connection in alternative 1(also flagged inline)? - I think I sent out the query to the implementors. I'll resend to the list. agree with 3 4,5 on a different thread 6 as you say, it depends on the implementations; I always like failures such as Event 17 to take us to Idle so I hope the answer is alternative 2 7 I am still struggling with; we are in OpenConfirm so we already have received an Open from the remote peer and Event 18 is a second Open from the same peer. Perhaps my struggle is that I think in terms of two (or more) FSM for a given IP address pair so the Open Collision detection will occur when the/an- other FSM receives a valid Open in states Active/Connect/Open Sent and will generate Event 22 into this FSM so Event 18 cannot occur. But yes, if Event 18 can occur in this FSM and this connection is to be dumped, then Idle state it should be as you suggest. I have slotted in [optionally] in front of the peer oscillation damping in your text because I think it should be optional:-) Collection collision is simple until I think about it :-( Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Susan Hares <shares@nexthop.com> To: Tom Petch <nwnetworks@dial.pipex.com>; Jeffrey Haas <jhaas@nexthop.com> Cc: idr <idr@merit.edu> Date: 05 December 2002 02:17 Subject: RE: bgp18 WG Last Call fsm missing next state Tom and Jeff: Thanks for the great review of the state machine. Please confirm I've got my 7 changes numbered right. Here's where we have consensus: change 1 - consensus, editorial state: Connect Events: 15 and 16 change 2 - no consensus I've given two alternatives. We need to get implementers to pick 1. thread: State: Connect Event: 14 change 3 - consensus, editorial state: Active: Events: 15 and 16 change 4 and 5: State: Active Events: 13 and 14 On changes 4 and 5 - moved to a different thread thread titled: fsm missing next state - Events 13 - 17 and the TCP connection on change 6 - no consensus State: Active Event: 17 Review my explanation and the new text, and see if we have consensus. on change 7: no consensus. State: Open Confirm Event: 18 Review my explanation and see if we have consensus. Sue Hares (skh@nexthop.com) ============= Warning - put the screen on wide mode for the following text: ------------------- Here's my response on the 7 changes 1) Connect State - Event 15 or Event 16 [consensus, editorial] note: The delay retry timer is utilized instead of the connect retry timer for the next two changes. Previous text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the "Delay Open Flag". If the delay open flag is set, the local system: - clears the connect retry timer, - sets the BGP Open delay timer to initial value If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clear BGP Open Delay timer (set to zero), - completes the BGP initialization, - send an Open message to its peer, - sets hold timer to a large value, and - change the state to Open Sent. New text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the Delay Open flag prior to processing: If the Delay Open flag is set, the local system: - clears the connect retry timer, - sets the BGP** Open Delay timer to initial value, and - stays in the Connect state. If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clears the BGP**Open Delay timer (sets to zero), - completes the BGP initialization, - sends an Open message to its peer, - sets the hold timer to a large value, and - changes the state to Open Sent. **tp agreed with changes to make it Open Delay timer 2)Connect State - Event 14 [no consensus] Current Text: If the TCP connection receives an indication that that is invalid or unconfigured [Event 14], - the TCP connection is rejected. At the very least this section needs more "word smithing", so I'd like to change it for more clarity at least. I'm not sure this represents the implementations. What I'd like to do is query the implementations to see what they do if they receive a valid TCP connection with an invalid or unconfigured peer. Two options: Alternative 1: Count it as a valid response New Text: If a TCP connection is received that has an invalid format, or an unconfigured host [Event 14], the local system: - rejects the TCP connection, - increments the connect retry counter, - performs bgp peer oscillation checks. If bgp peer oscillation checks allow for a new connection, the bgp peer - restarts the Connect retry timer with configured value, and - enters the Active state. **tp need a then clause here 'if bgp peer oscillation damping does not allow for a new connection, then the local system ???' FSM table: Idle Connect Active Open-Sent Open-Confirm Establish Event-14 ======================================================= Next state Idle | Active|Active|Open-Sent|Open-Confirm|Establish| ======================================================= action V | Y2 | L | Ignore | Ignore | Ignore | ======================================================= Alternative 2: Reject the connection and see if valid or configured one appears within the connect timer window. New Text: If a TCP connection is received that has an invalid format, or an unconfigured host [Event 14], the local system: - rejects the TCP connection, - and stays in the Connect state. FSM table: Idle Connect Active Open-Sent Open-Confirm Establish Event-14 ======================================================= Nxt state Idle |Connect|Active|Open-Sent|Open-Confirm|Establish| ======================================================= action V | L | L | Ignore | Ignore | Ignore | ======================================================= 3) Active State - Event 15 or Event 16 [consensus, editorial] Previous text: A TCP connection succeeds [Event 15 or Event 16], the local system: process the TCP connection flags - If the BGP delay open flag is set: - clears the connect retry timer [through the following text: - and changes its state to Open Sent. New text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the "Delay Open Flag" prior to processing: If the delay open flag is set, the local system: - clears the connect retry timer, - sets the BGP open delay timer to initial value, and - stays in the Active state. If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clears the BGP Delay timer (sets to zero), - completes the BGP initialization, - sends an Open message to its peer, - sets the hold timer to a large value, and - changes the state to Open Sent. 4) Active State, Event 13 [no consensus] 5) Active State, Event 14 [no consensus] The problem with this state is it is difficult to exactly specify without discussing the TCP Messages that FSM document covers. I'll query if the implementors require all of events 13-17 as mandatory. We'll 6) Active State, Event 17 [no consensus] comment: In the Active state, we may already have a connection and be awaiting the Open Delay timer. The TCP disconnect or timeout could occur in this state due to the "Open Delay Timer". If the TCP Disconnect is ignored, we could have some peer oscillation. If the we wait, then the connection retry timer needs to be kept running. The text below allows this timer. The real question is what is the status of the current implementations. I agree, the Active state and the connect state should match. Old Text: If the TCP connection fails (timeout or disconnection) [Event 17], the local system: - set TCP disconnect in the MIB reason code, - restart Connect retry timer (with initial value), - release all BGP resources, - Acknowledge the Drop of the TCP connection if TCP disconnect (FIN ACK), - Increment ConnectRetryCnt (connect retry count) by 1, and - performs the BGP peer oscillation damping process. Applicable FSM State table: FSM table old: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Active |Idle |Active | Idle |Idle | | | | | | | ========================================================= action V | Y2 | G | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= Alternative 1: FSM table new: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ======================================================== Next state Idle |Active |Active |Active | Idle |Idle | | | | | | | | ============================================================== action V | G | G | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| =============================================================== G: The local system: - restarts the connect retry timer (at intial value), - continues to listen for a connection that may be initiated by the remote peer, and - sets its next state to Active. New Text: (for Connect and Active state) If the TCP connection fails (timeout or disconnect) [Event 17], the local system: - restarts the connect retry timer, - continues to listen for a connection that may be initiated by the remote BGP peer, and - changes it state to Active. Alternative 2: FSM table new: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Idle |Idle |Active | Idle |Idle | | | | | | | ========================================================= action V | Y2 | Y2 | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= Next Text: If the location system receives a TCP connection failed [Event 17], the local system will: - increment the ConnectRetrycnt (connect retry count) by 1, - release all BGP resources associated with this connection, - perform BGP peer oscillation (if configured), and - go to Idle Y2 - is: The local system: 1) increments the ConnectRetryCnt (connect retry count) by 1, 2) releases all BGP resources associated with this connection, and 3) performs the BGP peer oscillation damping process if the damping process allows for a new connection, the local system - restarts the connect retry timer (with initial value, and - goes to Idle If the damping process does not allow for a new connection, the local system - set the flags to damp the creation of a new bgp connection until a manual start occurs, and - goes to Idle. 7) Open Confirm state, Event 18 Here's my proposed text. Please let me know what you think. I think this is an editorial change. Old text: If the open message is valid, the collision detect function is processed per section 6.8. If this connection is to be dropped due to call collision, the local system: - sends a Notification with a Cease - resets the Connect timer (to zero), - releases all BGP resources, - Drop the TCP connection (sends a TCP FIN), - increments the ConnectRetryCnt by 1 (connect retry count), and - performs an BGP peer oscillation damping process. New text: If the open message is valid, the BGP peer connection is check to detect a collision per section 6.8. If this connection is to be dropped due to call collision, the local system: - sends a Notification with a Cease - resets the Connect timer (to zero), - releases all BGP resources, - Drop the TCP connection (sends a TCP FIN), - increments the ConnectRetryCnt by 1 (connect retry count), and - **[optionally] performs an BGP peer oscillation damping processing, and - enters the Idle State. notes: Collision detect impacts Open Sent, Open Confirm, and Established states. Applicable FSM state >> Open Confirm State: >> >> If the Open messages is valid [Event 18], the collision >> detect function is processed per section 6.8. If this >> connection is to be dropped due to call collision, the >> local system: >>** enters what state > > This instance of the FSM should be discarded since it implies there is > another FSM instance running that will proceed to Established? -----Original Message----- From: Tom Petch [mailto:nwnetworks@dial.pipex.com] Sent: Monday, December 02, 2002 11:07 AM To: Jeffrey Haas Cc: idr Subject: Re: bgp18 WG Last Call fsm missing next state Jeff I appreciate your comments, agreeing with many. A minor supplementary; Open Confirm, valid open [event 18], decide to drop this connection and so the FSM but what about the actions of incrementing the ConnectRetryCnt (and optional peer oscillation damping)? Are we doing that in the other FSM? on a per router id or per IP address pair independent of FSM? And a slightly bigger issue in the Active state (which is the passive state because we have entered it from Idle with Passive TCP flag set Events 4/5). In which case, TCP Indication valid [event 13] or invalid [event 14] are what I would expect with event 14 keeping us in Active state; and if event 13 also keeps us in Active state (while sending a TCP SYN-ACK) then TCP Connection Confirmed [event 16] will ensue while TCP Connection Acknowledged [event 15] would be an FSM error. Except that you suggest that event 14 in the Connect state takes us to Active state which seems fine but for the fact that we will have sent a TCP SYN (because that is what the Connect state is), this invalid Indication may be a coincidence, in which case our SYN is about to receive a SYN-ACK back ie event 15 is valid after all. And I am then in agreement with you over events 15/16 in the Active (ie passive) state (delay open flag keeps us in the Active state). Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Jeffrey Haas <jhaas@nexthop.com> To: Tom Petch <nwnetworks@dial.pipex.com> Cc: idr <idr@merit.edu> Date: 30 November 2002 01:33 Subject: Re: bgp18 WG Last Call fsm missing next state >Tom, > >On Thu, Nov 28, 2002 at 02:53:56PM -0000, Tom Petch wrote: >> In seven places in the FSM, the actions listed for the occurrence of >> an event do not include any indication of what the next state should >> be, nor can I be sure I can guess what it should be. >> >> I believe the FSM should specify the new state in the cases listed >> below. >> Change 1: >> Connect State: >> >> If the TCP connection succeeds [Event 15 or >> Event 16], the local system checks the "Delay Open >> Flag". If the delay Open flag is set, the local system: >> **enters what state > >Remains in the connect state awaiting either an open message or >for the opendelay timer to expire. Agreed Change 2: > >> If the TCP connection receives an indication >> that is invalid or unconfigured. [Event 14]: >> **enters what state > >Enters the active state. >Connectretrytimer should be started. Agreed Change 3: >> Active State: >> >> A TCP connection succeeds [Event 15 or Event 16], the >> local system: process the TCP connection flags >> - If the BGP delay open flag is set: >> ** enters what state (I think this is an FSM error in TCP because it >> has not initiated a connection!) >In the active state, we are awaiting incoming tcp connections to be >completed. Here we should: > >remain in the active state >wait for an open message or the opendelay timer to expire. > Change 4: >> If the local system receives a valid TCP Indication >> [Event 13], the local system processes the TCP connection >> flags. >> ** enters what state > >I'm also confused by this event. Since we are not initiating a TCP >connection, but are awaiting one to complete, why would we get an >Indication here? FSM error? We would get an indication from the remote side that it had send us a TCP Syn. The real question is Change 5: >> If the local system receives a TCP indication >> that is invalid for this connection [Event 14]: >> ** enters what state > >Ditto. Change 6: >> >> If the local system receives a TCP connection >> failed [Event 17] (timeout or receives connection >> disconnect), the local system will: >> ** enters what state > >Ditto. Change 7: >> Open Confirm State: >> >> If the Open messages is valid [Event 18], the collision >> detect function is processed per section 6.8. If this >> connection is to be dropped due to call collision, the >> local system: >> ** enters what state > >This instance of the FSM should be discarded since it implies there is >another FSM instance running that will proceed to Established? > >> Tom Petch >> nwnetworks@dial.pipex.com > >-- >Jeff Haas >NextHop Technologies Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id RAA25840 for <idr-archive@nic.merit.edu>; Thu, 5 Dec 2002 17:18:32 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id D7161912D5; Thu, 5 Dec 2002 17:18:02 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 9E5AD912D7; Thu, 5 Dec 2002 17:18:02 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 622AE912D5 for <idr@trapdoor.merit.edu>; Thu, 5 Dec 2002 17:17:20 -0500 (EST) Received: by segue.merit.edu (Postfix) id 422C35DDFE; Thu, 5 Dec 2002 17:17:20 -0500 (EST) Delivered-To: idr@merit.edu Received: from nemesis.systems.pipex.net (nemesis.systems.pipex.net [62.241.160.8]) by segue.merit.edu (Postfix) with ESMTP id 5614D5DEAD for <idr@merit.edu>; Thu, 5 Dec 2002 17:17:19 -0500 (EST) Received: from tom3 (usermh96.uk.uudial.com [62.188.122.181]) by nemesis.systems.pipex.net (Postfix) with SMTP id AB7BA1600277B; Thu, 5 Dec 2002 22:17:09 +0000 (GMT) Message-ID: <018f01c29cac$089cd8c0$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: "Susan Hares" <shares@nexthop.com>, "Jeffrey Haas" <jhaas@nexthop.com> Cc: "idr" <idr@merit.edu> Subject: Re: bgp18 WG Last Call fsm missing next state Date: Thu, 5 Dec 2002 22:14:46 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Sue, Jeff Agree with 1 (with a slight change of words put **inline to make the name the Open Delay timer as Yakov accepted 2 wait on implementors - but do we need a then clause for when peer oscillation damping does not allow a new connection in alternative 1(also flagged inline)? agree with 3 4,5 on a different thread 6 as you say, it depends on the implementations; I always like failures such as Event 17 to take us to Idle so I hope the answer is alternative 2 7 I am still struggling with; we are in OpenConfirm so we already have received an Open from the remote peer and Event 18 is a second Open from the same peer. Perhaps my struggle is that I think in terms of two (or more) FSM for a given IP address pair so the Open Collision detection will occur when the/an- other FSM receives a valid Open in states Active/Connect/Open Sent and will generate Event 22 into this FSM so Event 18 cannot occur. But yes, if Event 18 can occur in this FSM and this connection is to be dumped, then Idle state it should be as you suggest. I have slotted in [optionally] in front of the peer oscillation damping in your text because I think it should be optional:-) Collection collision is simple until I think about it :-( Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Susan Hares <shares@nexthop.com> To: Tom Petch <nwnetworks@dial.pipex.com>; Jeffrey Haas <jhaas@nexthop.com> Cc: idr <idr@merit.edu> Date: 05 December 2002 02:17 Subject: RE: bgp18 WG Last Call fsm missing next state Tom and Jeff: Thanks for the great review of the state machine. Please confirm I've got my 7 changes numbered right. Here's where we have consensus: change 1 - consensus, editorial state: Connect Events: 15 and 16 change 2 - no consensus I've given two alternatives. We need to get implementers to pick 1. thread: State: Connect Event: 14 change 3 - consensus, editorial state: Active: Events: 15 and 16 change 4 and 5: State: Active Events: 13 and 14 On changes 4 and 5 - moved to a different thread thread titled: fsm missing next state - Events 13 - 17 and the TCP connection on change 6 - no consensus State: Active Event: 17 Review my explanation and the new text, and see if we have consensus. on change 7: no consensus. State: Open Confirm Event: 18 Review my explanation and see if we have consensus. Sue Hares (skh@nexthop.com) ============= Warning - put the screen on wide mode for the following text: ------------------- Here's my response on the 7 changes 1) Connect State - Event 15 or Event 16 [consensus, editorial] note: The delay retry timer is utilized instead of the connect retry timer for the next two changes. Previous text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the "Delay Open Flag". If the delay open flag is set, the local system: - clears the connect retry timer, - sets the BGP open delay timer to initial value If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clear BGP Open Delay timer (set to zero), - completes the BGP initialization, - send an Open message to its peer, - sets hold timer to a large value, and - change the state to Open Sent. New text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the Delay Open flag prior to processing: If the Delay Open flag is set, the local system: - clears the connect retry timer, - sets the BGP** Open Delay timer to initial value, and - stays in the Connect state. If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clears the BGP**Open Delay timer (sets to zero), - completes the BGP initialization, - sends an Open message to its peer, - sets the hold timer to a large value, and - changes the state to Open Sent. **tp agreed with changes to make it Open Delay timer 2)Connect State - Event 14 [no consensus] Current Text: If the TCP connection receives an indication that that is invalid or unconfigured [Event 14], - the TCP connection is rejected. At the very least this section needs more "word smithing", so I'd like to change it for more clarity at least. I'm not sure this represents the implementations. What I'd like to do is query the implementations to see what they do if they receive a valid TCP connection with an invalid or unconfigured peer. Two options: Alternative 1: Count it as a valid response New Text: If a TCP connection is received that has an invalid format, or an unconfigured host [Event 14], the local system: - rejects the TCP connection, - increments the connect retry counter, - performs bgp peer oscillation checks. If bgp peer oscillation checks allow for a new connection, the bgp peer - restarts the Connect retry timer with configured value, and - enters the Active state. **tp need a then clause here 'if bgp peer oscillation damping does not allow for a new connection, then the local system ???' FSM table: Idle Connect Active Open-Sent Open-Confirm Establish Event-14 ======================================================= Next state Idle | Active|Active|Open-Sent|Open-Confirm|Establish| ======================================================= action V | Y2 | L | Ignore | Ignore | Ignore | ======================================================= Alternative 2: Reject the connection and see if valid or configured one appears within the connect timer window. New Text: If a TCP connection is received that has an invalid format, or an unconfigured host [Event 14], the local system: - rejects the TCP connection, - and stays in the Connect state. FSM table: Idle Connect Active Open-Sent Open-Confirm Establish Event-14 ======================================================= Nxt state Idle |Connect|Active|Open-Sent|Open-Confirm|Establish| ======================================================= action V | L | L | Ignore | Ignore | Ignore | ======================================================= 3) Active State - Event 15 or Event 16 [consensus, editorial] Previous text: A TCP connection succeeds [Event 15 or Event 16], the local system: process the TCP connection flags - If the BGP delay open flag is set: - clears the connect retry timer [through the following text: - and changes its state to Open Sent. New text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the "Delay Open Flag" prior to processing: If the delay open flag is set, the local system: - clears the connect retry timer, - sets the BGP open delay timer to initial value, and - stays in the Active state. If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clears the BGP Delay timer (sets to zero), - completes the BGP initialization, - sends an Open message to its peer, - sets the hold timer to a large value, and - changes the state to Open Sent. 4) Active State, Event 13 [no consensus] 5) Active State, Event 14 [no consensus] The problem with this state is it is difficult to exactly specify without discussing the TCP Messages that FSM document covers. I'll query if the implementors require all of events 13-17 as mandatory. We'll 6) Active State, Event 17 [no consensus] comment: In the Active state, we may already have a connection and be awaiting the Open Delay timer. The TCP disconnect or timeout could occur in this state due to the "Open Delay Timer". If the TCP Disconnect is ignored, we could have some peer oscillation. If the we wait, then the connection retry timer needs to be kept running. The text below allows this timer. The real question is what is the status of the current implementations. I agree, the Active state and the connect state should match. Old Text: If the TCP connection fails (timeout or disconnection) [Event 17], the local system: - set TCP disconnect in the MIB reason code, - restart Connect retry timer (with initial value), - release all BGP resources, - Acknowledge the Drop of the TCP connection if TCP disconnect (FIN ACK), - Increment ConnectRetryCnt (connect retry count) by 1, and - performs the BGP peer oscillation damping process. Applicable FSM State table: FSM table old: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Active |Idle |Active | Idle |Idle | | | | | | | ========================================================= action V | Y2 | G | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= Alternative 1: FSM table new: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Active |Active |Active | Idle |Idle | | | | | | | ========================================================= action V | G | G | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= G: The local system: - restarts the connect retry timer (at intial value), - continues to listen for a connection that may be initiated by the remote peer, and - sets its next state to Active. New Text: (for Connect and Active state) If the TCP connection fails (timeout or disconnect) [Event 17], the local system: - restarts the connect retry timer, - continues to listen for a connection that may be initiated by the remote BGP peer, and - changes it state to Active. Alternative 2: FSM table new: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Idle |Idle |Active | Idle |Idle | | | | | | | ========================================================= action V | Y2 | Y2 | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= Next Text: If the location system receives a TCP connection failed [Event 17], the local system will: - increment the ConnectRetrycnt (connect retry count) by 1, - release all BGP resources associated with this connection, - perform BGP peer oscillation (if configured), and - go to Idle Y2 - is: The local system: 1) increments the ConnectRetryCnt (connect retry count) by 1, 2) releases all BGP resources associated with this connection, and 3) performs the BGP peer oscillation damping process if the damping process allows for a new connection, the local system - restarts the connect retry timer (with initial value, and - goes to Idle If the damping process does not allow for a new connection, the local system - set the flags to damp the creation of a new bgp connection until a manual start occurs, and - goes to Idle. 7) Open Confirm state, Event 18 Here's my proposed text. Please let me know what you think. I think this is an editorial change. Old text: If the open message is valid, the collision detect function is processed per section 6.8. If this connection is to be dropped due to call collision, the local system: - sends a Notification with a Cease - resets the Connect timer (to zero), - releases all BGP resources, - Drop the TCP connection (sends a TCP FIN), - increments the ConnectRetryCnt by 1 (connect retry count), and - performs an BGP peer oscillation damping process. New text: If the open message is valid, the BGP peer connection is check to detect a collision per section 6.8. If this connection is to be dropped due to call collision, the local system: - sends a Notification with a Cease - resets the Connect timer (to zero), - releases all BGP resources, - Drop the TCP connection (sends a TCP FIN), - increments the ConnectRetryCnt by 1 (connect retry count), and - **[optionally] performs an BGP peer oscillation damping processing, and - enters the Idle State. notes: Collision detect impacts Open Sent, Open Confirm, and Established states. Applicable FSM state >> Open Confirm State: >> >> If the Open messages is valid [Event 18], the collision >> detect function is processed per section 6.8. If this >> connection is to be dropped due to call collision, the >> local system: >>** enters what state > > This instance of the FSM should be discarded since it implies there is > another FSM instance running that will proceed to Established? -----Original Message----- From: Tom Petch [mailto:nwnetworks@dial.pipex.com] Sent: Monday, December 02, 2002 11:07 AM To: Jeffrey Haas Cc: idr Subject: Re: bgp18 WG Last Call fsm missing next state Jeff I appreciate your comments, agreeing with many. A minor supplementary; Open Confirm, valid open [event 18], decide to drop this connection and so the FSM but what about the actions of incrementing the ConnectRetryCnt (and optional peer oscillation damping)? Are we doing that in the other FSM? on a per router id or per IP address pair independent of FSM? And a slightly bigger issue in the Active state (which is the passive state because we have entered it from Idle with Passive TCP flag set Events 4/5). In which case, TCP Indication valid [event 13] or invalid [event 14] are what I would expect with event 14 keeping us in Active state; and if event 13 also keeps us in Active state (while sending a TCP SYN-ACK) then TCP Connection Confirmed [event 16] will ensue while TCP Connection Acknowledged [event 15] would be an FSM error. Except that you suggest that event 14 in the Connect state takes us to Active state which seems fine but for the fact that we will have sent a TCP SYN (because that is what the Connect state is), this invalid Indication may be a coincidence, in which case our SYN is about to receive a SYN-ACK back ie event 15 is valid after all. And I am then in agreement with you over events 15/16 in the Active (ie passive) state (delay open flag keeps us in the Active state). Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Jeffrey Haas <jhaas@nexthop.com> To: Tom Petch <nwnetworks@dial.pipex.com> Cc: idr <idr@merit.edu> Date: 30 November 2002 01:33 Subject: Re: bgp18 WG Last Call fsm missing next state >Tom, > >On Thu, Nov 28, 2002 at 02:53:56PM -0000, Tom Petch wrote: >> In seven places in the FSM, the actions listed for the occurrence of >> an event do not include any indication of what the next state should >> be, nor can I be sure I can guess what it should be. >> >> I believe the FSM should specify the new state in the cases listed >> below. >> Change 1: >> Connect State: >> >> If the TCP connection succeeds [Event 15 or >> Event 16], the local system checks the "Delay Open >> Flag". If the delay Open flag is set, the local system: >> **enters what state > >Remains in the connect state awaiting either an open message or >for the opendelay timer to expire. Agreed Change 2: > >> If the TCP connection receives an indication >> that is invalid or unconfigured. [Event 14]: >> **enters what state > >Enters the active state. >Connectretrytimer should be started. Agreed Change 3: >> Active State: >> >> A TCP connection succeeds [Event 15 or Event 16], the >> local system: process the TCP connection flags >> - If the BGP delay open flag is set: >> ** enters what state (I think this is an FSM error in TCP because it >> has not initiated a connection!) >In the active state, we are awaiting incoming tcp connections to be >completed. Here we should: > >remain in the active state >wait for an open message or the opendelay timer to expire. > Change 4: >> If the local system receives a valid TCP Indication >> [Event 13], the local system processes the TCP connection >> flags. >> ** enters what state > >I'm also confused by this event. Since we are not initiating a TCP >connection, but are awaiting one to complete, why would we get an >Indication here? FSM error? We would get an indication from the remote side that it had send us a TCP Syn. The real question is Change 5: >> If the local system receives a TCP indication >> that is invalid for this connection [Event 14]: >> ** enters what state > >Ditto. Change 6: >> >> If the local system receives a TCP connection >> failed [Event 17] (timeout or receives connection >> disconnect), the local system will: >> ** enters what state > >Ditto. Change 7: >> Open Confirm State: >> >> If the Open messages is valid [Event 18], the collision >> detect function is processed per section 6.8. If this >> connection is to be dropped due to call collision, the >> local system: >> ** enters what state > >This instance of the FSM should be discarded since it implies there is >another FSM instance running that will proceed to Established? > >> Tom Petch >> nwnetworks@dial.pipex.com > >-- >Jeff Haas >NextHop Technologies Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id QAA23172 for <idr-archive@nic.merit.edu>; Thu, 5 Dec 2002 16:27:48 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 8B7FD912D4; Thu, 5 Dec 2002 16:27:20 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 54C4A912D5; Thu, 5 Dec 2002 16:27:20 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 9AD18912D4 for <idr@trapdoor.merit.edu>; Thu, 5 Dec 2002 16:27:18 -0500 (EST) Received: by segue.merit.edu (Postfix) id 7E27D5DF40; Thu, 5 Dec 2002 16:27:18 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id B6AE95DF3F for <idr@merit.edu>; Thu, 5 Dec 2002 16:27:17 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB5LRGg66961 for idr@merit.edu; Thu, 5 Dec 2002 16:27:16 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB5LR7C66941 for <idr@merit.edu>; Thu, 5 Dec 2002 16:27:07 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message Subject: Comments 11-20 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 Date: Thu, 5 Dec 2002 16:27:07 -0500 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AB0B@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: Comments 11-20 Thread-Index: AcKcpRAd4EMRRevxTaq5YxZSayJvCg== From: "Susan Hares" <shares@nexthop.com> To: "Yakov Rekhter" <yakov@juniper.net>, "Jeffrey Haas" <jhaas@nexthop.com>, <siva@ctd.haltech.com> Cc: <idr@merit.edu> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id QAA23172 Siva Again, thanks for all the comments. Here's the summary of 11-20 comment: 11 - no consensus, no text supplied note: usage varies from my understanding of common usage. 12 - consensus, no text supplied see comment 9 and proposed text below. 13a - no consensus, see alternate text 13b - no consensus, implementation specific detail 14 - no consensus, disagree with comment policy issues, outside scope of document 15 - no consensus, see alternative text. 16 - no consensus, see my proposed text. You did not propose text See alternate text for events 13-17 Also note, that I consider that events 13 and 14 should go to optional. 17 - consensus, agree to change in text. 18 - agree needs to be fixed, see my alternative text. 19 - further clarity needed, please respond to question 20 - agree, text will be changed So, review the comments and let me know. Comments 21-30 in the next message. Thanks for working so hard on the comments. Sue ===================== 11) Event 10 (Hold timer) We use the hold timer for two purposes * Waiting for an open message (with a default value of 240 seconds) * Waiting for Keepalives (with a default value of 90 seconds) Can we use two different timers (or at least call them two different timer events) ? Siva: I do not believe that most implementatios utilize two timers. Can you indicate implementations that do? ======================================================================= 12) Event 12 > Event12: DelayBGP Open timer expires [changed] > > Definition: A timer that delays sending of the BGP > Open message for n seconds after the > Transport connection has been completed.= >The Definition seems to better define what the timer is, rather than = > what > its expiry means. Can we change it to reflect this ? new text: see comment on 9 on Timer definition, and earlier change. Event 12: An event generated when the Delay BGP Open timer expires. ========================================= 13) Event 13, 14 definition > Event13: Transport Connection Indication & valid remote peer[changed] > > Definition: Event indicating that TCP connection > request with a valid source IP address and TCP > port, and valid destination IP address > and TCP Port. The definition of > invalid Source, and invalid Destination= > IP address is left to the implementation. a) How about changing this slighty to: % Definition: Event indicating receiving a TCP connection % request with a valid source IP address and TCP % port, and valid destination IP address= % and TCP Port. The definition of invalid My completion of your sentence: source, and invalid destination IP address is left to the implementation. How about: Definition: Event indicating the reception of a TCP connection request with a valid source IP address and TCP port, and valid destination IP address and TCP Port. The definition of invalid source address and port and invalid destination address is left to the implementation. Please comment if you agree with this change. b) Is it correct that Destination IP address validation is used when we do not wish to accept BGP connections on one of the local IP addresses (either because we wish to accept such addresses on only our loopback interfaces, or because the address to which we received the connection was not the optimal interface with which to speak to the sender) ? If so, can we add this to the description ? No, this is an implementation dependent change. ================================================================================= 14) Event 13,14 > Event13: Transport Connection Indication & valid remote peer[changed] > > Definition: Event indicating that transport > Request valid source IP address and TCP=20 > port, and valid destination IP address=20 > and TCP Port. The definition of=20 > invalid Source, and invalid Destination=20 > IP address is left to the implementation. > > a) I am unclear as to what validation of source and destination port = > mean. > > The source port's value should be of no concern to connection = > receipient system. The definition of "valid" is an implementation policy issue. Validating is the process of checking to make sure it is valid for the implementation's policy. > The destination port cannot be invalid, since BGP would not receive the > connection request in such a situation. Again, this is a policy issue. We are providing hooks for policy to be engaged. Policy is outside the scope of this document. > b) I think it will be useful to state what we consider to be an invalid > IP address. Again, it is policy issue which is implementation dependent. ==================== 15) Event 17 > Event17: TCP connection fails > This event can occur even when the transport connection is closed by > the other end. Since this does not reflect a 'failure ', can we change > the event name to > > % Event17: TCP connection termination Discussion: It both terminates from the remote site and can "timeout" - fail. Suggestions? I can use "disconnect", what do you think. ========================== 16) Events related to BGP Messages Can we make the definition style across these events consistent ? thanks for the comment. proposed text below, please agree on it. proposed text replacement ------------------------ Event13: TCP connection indication and valid remote peer Definition: Event indicating the local system reception of a TCP connection request with a valid source IP address and TCP port, and valid destination IP address and TCP Port. The definition of invalid source, and invalid destination IP address is left to the implementation. BGP's destination port SHOULD be port 179 as defined by IANA. TCP connection request is denoted by the local system receiving a TCP SYN. Status: Mandatory (Optional) Event14: RCV TCP connection indication with invalid source or destination Definition: Event indicating the local system reception of a TCP connection request with either an invalid source address or port number or an invalid destination address or port number. BGP destination port number SHOULD be 179 as defined by IANA. Again, a TCP connection request denoted by local system receiving a TCP SYN. Status: Mandatory (Optional) Event15: TCP connection request sent received an ACK. Definition: Event indicating the Local system's request to establish a TCP connection to the remote peer. The local system's TCP session sent a TCP SYN, and received a TCP SYN, ACK pair of messages, and Sent a TCP ACK. Status: Mandatory Event16: TCP connection confirmed Definition: Event indicates that the local system receiving a confirmation that the TCP connection has been established by the remote site. The remote peer's TCP engine sent a TCP SYN. The local peer's TCP engine sent a SYN, ACK pair, and now has received a final ACK. Status: Mandatory Event17: TCP connection fails Definition: Event indicates that the local system has received a TCP connection failure notice. The remote BGP peer's TCP machine could have sent a FIN. The local peer would respond with a FIN-ACK. Another alternative is that the local peer indicated a timeout in the TCP session and downed the connection. Status: Mandatory --------------------- 17) Event 19 Definition > Event19: BGPOpen with BGP Delay Open Timer running [changed] > > Definition: An event indicating that a valid Open=20 > Message has been successful=20 > established for a peer that is=20 > currently delaying the sending of an=20 > BGP Open message.=20 The definition appears a little fragmented. How about: % Definition: An event indicating that a valid OPEN % Message has been received for a peer % that has a successfully established % transport connection and is currently % delaying the seending of a BGP open % message agreed, changed in text. I will send out definition text for review in a separate message. ============================================================== 18) Event 22 Definition a)=20 > Event22: Open collision discard > > Definition: An event generated administratively=20 > when a connection Collision has been=20 > detected while processing an incoming =20 > Open message. This connection has been=20 Isn't this event 'automatically' generated, since it is a system generated event ?=20 response: How this generated is implementation specific. The "administratively" is to cover policy. b) > Event 22 is an administrative could=20 > occur if FSM is implemented as two=20 > The word event is missing. How about % Event 22 is an automatic event that % could occur if FSM is implemented as two=20 I've re-written the whole definition. What do you think about: Event22: Open collision dump Definition: An event generated administratively when a connection collision has been detected while processing an incoming OPEN message and this connection has been selected to disconnected. See Section 6.8 for more information on collision detection. Event22 is an administrative based only implementation specific policy. This Event may occur if the FSM is implemented as two linked state machines. Please confirm that you like this. ==================================== 19) FSM Description: ConnectRetry Count > The Connect Retry count is updated by the FSM but never used. In the > absence of peer oscillation damping, will this be used to stop > connection establishment attempts after a certain maximum number ? Yes, this is either implementation specific or is it based on the peer oscillation damping draft. > Can we include the use of this counter in some place ? Connect retry counter 1) Will be utilized by the peer oscillation damping drat. 2) Will be included in bgp-4-mibv2-xx. I just check and I didn't find it. Do you still want text in the main ==================================== 20) Handling Event 7 (Auto Stop) to Idle State processing: The handling of Event 7 is missing from the Idle State processing. Can we add this ? How about replacing > An manual stop event (Event2) is ignored in the Idle state. with % Manual stop (Event 2) and Auto stop (Event 7) events are ignored % in the Idle state Good catch, we'll add that in the text. Sorry it got dropped on the last revision. Agreed ============================== Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id QAA22812 for <idr-archive@nic.merit.edu>; Thu, 5 Dec 2002 16:21:35 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id E29EB912C6; Thu, 5 Dec 2002 16:21:13 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id A9CBC912D4; Thu, 5 Dec 2002 16:21:13 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 68243912C6 for <idr@trapdoor.merit.edu>; Thu, 5 Dec 2002 16:21:11 -0500 (EST) Received: by segue.merit.edu (Postfix) id 537F25DF13; Thu, 5 Dec 2002 16:21:11 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id D8B335DEF6 for <idr@merit.edu>; Thu, 5 Dec 2002 16:21:10 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB5LL9J66628 for idr@merit.edu; Thu, 5 Dec 2002 16:21:09 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB5LL0C66608 for <idr@merit.edu>; Thu, 5 Dec 2002 16:21:00 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message MIME-Version: 1.0 Subject: Content-Type: text/plain; charset="iso-8859-1" X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 Date: Thu, 5 Dec 2002 16:21:00 -0500 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AB0C@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Index: AcKcpDUtR520mUFcRNeIjnoraPoqqA== From: "Susan Hares" <shares@nexthop.com> To: "Yakov Rekhter" <yakov@juniper.net>, "Jeffrey Haas" <jhaas@nexthop.com>, <siva@ctd.haltech.com> Cc: <idr@merit.edu> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id QAA22812 Siva: This is my response to your comments 21-30. Thanks again for all the comments. Your review, Tom Petches, and Jeff's may help tighten the text. Please review my response. Let's see what we can get further consensus on. thanks for your hard work, Sue =================== Summary: 21 - no consensus, see questions 22 - no consensus, suggest folding this comment into discussion on events 13-17 on list 23 - no consensus, no text 24 - no consensus, alternative text suggested 25 - move to own mail thread 26 - move to mail thread on events 13-17 27 - move to mail thread on events 13-17 28 - no consensus, but close good catch on the 4 minutes, Please choose one of the alternative texts. 29 -no consensus, see my alternative text 30 -no consensus mail list indicated no routes in active state. =============================== 21) Clearing of Connection Retry Timer: There are a few sections where the FSM draft states that the Connection Retry timer needs to be reset, wheras the conenct retry timer had been cleared prior to entering that state. We can remove these instructions to clear the connect retry timer. List of places where the connect retry timer need not be cleared a) Handling of Event 19 in the Connect State b) Handling of Events 12 in the Active State c) All cases where it is refered to in the OpenSent, OpenConfirm and Established states Comment: 1) Does it hurt to have the connect retry timer cleared at these points, since it has already been cleared. I felt it eased the implementations to allow the action routines to be shared across as many states as possible. You can see this a bit more actively. ==================== 22) Handling of Event 14 in the Connect state > If the transport connection receives an indication > that is invalid or unconfigured. [Event 14]: > - the TCP connection is rejected. > > I don't understand how we would get this event while in this state. See my earlier comments (1-10) on the connection state. It happens in implementations which track the TCP state more closely. I suggest that Event 14 become optional. =============================================================== 23) Handling events 20, 21 in the Connect State and Active State We need to consider the case where we receive events 20 (message header error) and 21 (Open message error) when the delay timer is running. Since the connection has been established at this point, we need to send a Notification message and then terminate the connection. Alternative comments: 1) We have not sent an Open statement. 2) Why do we have to send an Notification? I see no justification for it. Suggestion: Do you have implementations that send notification? Do you know of others that don't. =========================================================== 24) Handling the default events in the Connect state The BGP Delay OpenTimers needs to be cleared if it is running. >How about adding this: > > % - If the ConnectRetry Timer is running > % - Clear the Connect Retry timer > % - Otherwise > % - Clear the BGP Delay Open Timer response: By the default you mean the text: In response to any other events[Events 7-8, 10-11, 18, 20-27], the local system: "resets" to me implies stops and clears. I think the text is clear than the text above. ------------ Is this the replacement text you imply above: - resets the connect retry timer (sets to zero), - clears the BGP Delay timer (sets to zero), - increments the ConnectRetryCnt (connect retry count) by 1, - [optionally] performs bgp peer oscillation damping, and - goes to Idle text: ================================================ 25) Handling Event 23 (Notify with version mismatch error) in Connect and OpenSent >This is currently being handled in the default event processing = > section. > However, we do not need to go through the peer oscillation damping > process in this case. Can we change the wordings to reflect this, or > move this out of peer oscillation damping processing ? response: 1) There is no default event handling process in the text, you will need to specify the text. 2) The state table below (hares-statemt-03.txt) states shows the changes ------------- Event 23 states: current Idle Connect Active Open-Sent Open-Cnf Establish ---------------------------------------------------- next state Idle Idle Idle Idle Idle Idle ----------------------------------------------- action V D D Y Y T ============================================== V - Indicate FSM errors and ignore. D - 1) resets the connect retry timer (sets to zero), 2) drops the TCP connection, 2) releases all BGP resources, 3) increments the ConnectRetryCnt (connect retry count) by 1, 4) [optionally] performs the bgp peer oscillation damping, and Goes to Idle state. Y 1) resets the connect retry timer (sets to zero), 2) Drops the TCP connection, 3) releases all BGP resources, 4) [optionally] ============================= 26) Handling Event 17 in The connect state > If the transport connection fails (timeout or transport > disconnect) [Event17], the local system: > - changes its state to Active.=20 >If the transport connection fails when the BGP Open Delay timer is > running, should we still be going into the Active state ? see mail thread on eventse 13-17 ========================================== 27) Handling Event 17 in Active state We should now move into Idle state. Can we add % - Goes to Idle state See mail thread on events 13-17, ===================================== 28) Handling Event 19 in the Active state > - Set the Hold timer to a large value (4 minutes), Since OPEN messages have been exchanged, can we change this to % - If the negotiated Hold time is not 0, set the Hold time to % - the negotiated value The text in Active and Open Sent needs to be the same. The text in Open Sent is: - sets the Hold timer according to the negotiated value (see section 4.2), and Which text do you prefer? ========================================= 29) Handling Event 2 in Active state >> A manual stop event[Event2], the local system: >> - Sends a notification with a Cease, >> - drops the Transport connection,=20 > >These two actions are possible only if a transport connection had = > already been established. How about changing the text to > > % - If a transport connection had been successfully established > % - Send a Notification with a Cease > % - Drop the Transport Connection How about this text: A manual stop event [Event 2], the local system - Drop the TCP connection, - Release all BGP resources, - resets the connection retry timer [sets to zero], - goes to Idle. ===================== 30) Default event handling in Active State To ensure consistencey with E2 handling, can we add % - If any BGP Routes exist, delete the routes Comment: Yakov and Jeff noted, there are no routes in Active state. =============== Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id QAA22680 for <idr-archive@nic.merit.edu>; Thu, 5 Dec 2002 16:18:58 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 92EF79123E; Thu, 5 Dec 2002 16:18:37 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 5EB5C912C6; Thu, 5 Dec 2002 16:18:37 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id DB12B9123E for <idr@trapdoor.merit.edu>; Thu, 5 Dec 2002 16:18:35 -0500 (EST) Received: by segue.merit.edu (Postfix) id BEA065DF2D; Thu, 5 Dec 2002 16:18:35 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 1EC4D5DF13 for <idr@merit.edu>; Thu, 5 Dec 2002 16:18:35 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB5LIUc66562 for idr@merit.edu; Thu, 5 Dec 2002 16:18:30 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB5LIIC66534 for <idr@merit.edu>; Thu, 5 Dec 2002 16:18:18 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message Subject: Comments 30-36 from Siva MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 Date: Thu, 5 Dec 2002 16:18:18 -0500 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AB0E@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: Comments 30-36 from Siva Thread-Index: AcKco9S0K1ESZJE/Tz29y+s/qzoqAA== From: "Susan Hares" <shares@nexthop.com> To: <siva@ctd.haltech.com>, <idr@merit.edu> Cc: <yakov@juniper.net>, <jhass@nexthop.com> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id QAA22680 Siva: thanks again for all the comments. Let's keep working until we get a few more of these into the consensus model. ======================= Here's the summary: 31 - no consensus, question issue resolved one way last Jan-March Clearing of Hold time included in Clear BGP resources. 32 - no consensus, issue resolve one way last Jan - March Clearing of keep alive timer included in Clear BGP resources 33 - no consensus, query implementations Exact Open Exchange is being debated. What exactly occurs in implementations? 34 - consensus to pull mibs. 35 - no consensus, text approach suggested 36 - no consensus, no desire to open issue [author/chair hat on] I tried to add new state in 1Q 2002. Don't want to go through another 3-4 months of debate. If purchasers of boxes w/BGP mandate, we'll reconsider. [author/chair hat off] Sue Hares =========================== 31) Clearing Hold timer in OpenSent, OpenConfirm and Established State > In all event handling where we go to Idle state, we need to clear the > Hold Timer as well. response: &Earlier comments (last January -March on mailing list), indicating that & & - Clearing the BGP resources & &included clearing the Hold timer as well. Do you disagree? =============== 32) Clearing Keepalive timer in OpenConfirm and Established State >In all event handling where we go to Idle state, we need to clear the >Keepalive Timer as well. response: & Again, earlier comments (Last January - March on mail list indicated that & clearing the BGP resources included this. Are you suggesting that & we break this out from BGP resources. [I used to have a number of these timers broken out.] ==================================== 33) Handling Event 18 in the OpenSent state >Why do we start the Keepalive timer at this stage ? Isn't it sufficient = >to do so when we move into Established state ? An earlier comment from Tom (and you) requested the following: <---Open [Open sent state] Open--> [Event 18] <---Open <---Keepalive [Action from Event 18 in Open Sent] [Open Confirm] Keepalive -> [Event 25] [established] What do implementations do? We'll have to query implementations. ============= 34) Established State > Some event handling in the Established state do not set the MIB Reason = >when > handling an event that causes an error. Can we add this ? we have pulled out all the MIB reason codes. ======================= General Comments ========================================== 35) For the events whose status is optional, can we state the impact of = not supporting them (in terms of any interoperability issues). I = understand that most of the optional events will not have such an impact; but a = clarification statement for the optional events would benefit new implementors. Response: Much of the support of optional parameters depends on policy. I could put a short note about the optional events and parameters as part of 8.1.5 or 8.2.1.3 I think it fits better in 8.1.5. Optional: Events: 3-8, 12, 13-14[my suggestion] 19, 22 Timers: Idle Hold Timer Open Delay Timer Required flags for optional parameters: Open Delay Flag BGP Stop Flap (I'm sure I've missed a few after 24 hours of non-stop comment resolution). If this is on the right track, I'll try to write something up. =========================== 36) Is delaying the sending of an OPEN message a standrad industry = practice ? > Also, in the FSM, this has been handled by practically implementing a > sub-state each, within the CONNECT and ACTIVE states. Won't the FSM > look more simple if we just had a new DelayOpen state that we could > move into ? [author hat on] ok, I tried the "new state" approach last January (2002), it went down to a resounding defeat. I'm not likely to try to this again. So, yes I proposed new states. No, it did not have consensus. No, I do not think it advisable to re-open thie issues. Whether we have sub-states or a new state, it will operate and function the same. The purist may like it more, but I really don't want to re-enter the 2-3 month debate on new states. [author hat off] Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id OAA15092 for <idr-archive@nic.merit.edu>; Thu, 5 Dec 2002 14:03:08 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 16ED7912D3; Thu, 5 Dec 2002 14:01:55 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 38096912D4; Thu, 5 Dec 2002 14:01:51 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 91FB6912D3 for <idr@trapdoor.merit.edu>; Thu, 5 Dec 2002 14:01:49 -0500 (EST) Received: by segue.merit.edu (Postfix) id 772625DE94; Thu, 5 Dec 2002 14:01:49 -0500 (EST) Delivered-To: idr@merit.edu Received: from nemesis.systems.pipex.net (nemesis.systems.pipex.net [62.241.160.8]) by segue.merit.edu (Postfix) with ESMTP id 121DF5DE14 for <idr@merit.edu>; Thu, 5 Dec 2002 14:01:49 -0500 (EST) Received: from tom3 (userar74.uk.uudial.com [62.188.136.233]) by nemesis.systems.pipex.net (Postfix) with SMTP id 87BE116007E87; Thu, 5 Dec 2002 19:01:41 +0000 (GMT) Message-ID: <016c01c29c90$ba0971c0$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: "Susan Hares" <shares@nexthop.com>, "idr" <idr@merit.edu> Cc: "Jeffrey Haas" <jhaas@nexthop.com> Subject: Re: bgp18 WG Last Call fsm missing keepalive Date: Thu, 5 Dec 2002 18:55:36 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Sue (I promise I won't mention that state again) 1) yes I think the same text in both Active and Connect states is a good resolution 2) complicated. As the fsm text stands, Event 19, along with a host of others, takes us back from Open Sent to Idle (I assume on the grounds this is an error condition) which seems very reasonable. But ...in quite a few places, such as Connect state events 2, 7,8,9,10,11, 17, 18, 20 thru 27, we do not stop the OpenDelay timer when going to Idle or Active so we could then go from eg Idle with Manual start [event 1] to Connect to Open Sent all before the OpenDelay timer expires in which case event 19 can occur validly in Open Sent - obscure but possible. (This is also true with Active state and events 2, 17 and the default list at the end). But I think this is an error, and that when exiting Connect state or Active state as listed above, we should have an additional action to stop the OpenDelay timer in which case event 19 in Open Sent becomes an error condition (again). But but but as ever, I cannot speak with authority for implementations and so if implementations do not stop the OpenDelay timer when exiting as above, then Event 19 is valid in Open Sent state - obscure but possible (again). My wish is to add the extra action, stop OpenDelay timer, for the events listed above in Active and Connect states in the expectation that that is what people have or should have implemented. Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Susan Hares <shares@nexthop.com> To: Tom Petch <nwnetworks@dial.pipex.com>; idr <idr@merit.edu> Cc: Jeffrey Haas <jhaas@nexthop.com> Date: 04 December 2002 19:50 Subject: RE: bgp18 WG Last Call fsm missing keepalive Tom: I agree about the putative OpenDelay, but I do not feel that represents the BGP implementations of today. We spent a long time Jan-Feb 2002 on adding any FSM States. My earlier note was too brief. Please answer questions 1) Agree, this can occur in the Active state as well as the Connect state. Will you accept the earlier text below to be inserted both places? Background: The state machine for Event 19 is: Idle Connect Active Open Open Estb Sent Confirm =============================================== Event 19 | | | | | | | next state |Idle | Open | Open | Open |Idle | Idle | | | confirm| confirm| Confirm| | | =============================================== action | V | N-2 | N-2 | N | E-1 | E | =============================================== Per the State Machine. Action v - FSM Error Action E - FSM Error, drop connection - etc, drop routes Action E-1 - FSM Error, drop connection (lots of Action N-2 (text below) Action N (text below, without sending Open) 2) Do you think that Event 19 is possible in the Open Sent state? Please answer this separately. Thanks for all the wonderful review of the State machine. I'm working my way through the comments. Sue -----Original Message----- From: Tom Petch [mailto:nwnetworks@dial.pipex.com] Sent: Wednesday, December 04, 2002 1:30 PM To: Susan Hares; idr Cc: Jeffrey Haas Subject: Re: bgp18 WG Last Call fsm missing keepalive Still not quite. This can occur in the Connect state as well as Active, when a TCP connection completes with delay open flag set, my putative OpenDelay state (sigh, point taken). And we should still go to OpenConfirm state at the bottom of the action list as is currently part of list H. Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Susan Hares <shares@nexthop.com> To: Tom Petch <nwnetworks@dial.pipex.com>; idr <idr@merit.edu> Cc: Jeffrey Haas <jhaas@nexthop.com> Date: 04 December 2002 18:07 Subject: RE: bgp18 WG Last Call fsm missing keepalive Tom: Just to clarify this comment: Event 19 = valid open with delay timer running **Connect or Active = 1) awaiting TCP connection, or 2) TCP connection completed and awaiting the TCP connection with delay timer running Case 1: - should not see Event 19 In transition from Active to Open Confirm, the connection must have a TCP connection completed. Case 1 does not have this occurring, so the transition must be avoided. Case 2: - should see Event 19 - Open, Keepalive should be sent. Previous text: (Action H from FSM document) If an Open is received with the BGP Delay Open timer running, [Event 19], the local system: - clears the connect retry timer [cleared to zero), - completes the BGP initialization, - stop and clears the BGP Open Delay timer, - Sends an Open Message, - sets the hold timer to a large value (4 minutes), and - changes its state to an Open Confirm. New text: [a New Action - N-2 : N + BGP keepalive sent] If an Open is received with the BGP Delay Open Timer running [Event 19], the local system: - **clears the connect retry timer [cleared to zero], - completes the BGP initialization, - stops and clears the BGP Open Delay timer, - **Sends an Open message, - Sends a Keepalive message, - If hold timer value is non-zero, - **sets keepalive timer - hold timer reset to negotiated value else if hold timer is zero, - **resets the keepalive timer, and - **resets the hold timer. ** and changes state to OpenConfirm - If the value of the autonomous system field is the same as the local Autonomous system number, set the connection status to a internal connection; otherwise it is "external". Please confirm that this change is OK for this comment. Sue -----Original Message----- From: Tom Petch [mailto:nwnetworks@dial.pipex.com] Sent: Thursday, November 28, 2002 9:54 AM To: idr Subject: bgp18 WG Last Call fsm missing keepalive When the OpenConfirm state is entered from OpenSent with the receipt of a valid open [Event 18], then a KeepAlive message is sent and the timer is started. When the OpenConfirm state is entered from Active or Connect on receipt of a valid open [Event 19], no message is sent, no timer is started. I believe this inconsistency is an error and should be corrected by adding these two actions in those two places. Tom Petch nwnetworks@dial.pipex.com Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id MAA08898 for <idr-archive@nic.merit.edu>; Thu, 5 Dec 2002 12:16:05 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id D84FD91207; Thu, 5 Dec 2002 12:15:15 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 479729120E; Thu, 5 Dec 2002 12:15:14 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id A29F791207 for <idr@trapdoor.merit.edu>; Thu, 5 Dec 2002 12:15:09 -0500 (EST) Received: by segue.merit.edu (Postfix) id D66185E135; Thu, 5 Dec 2002 12:15:08 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 50AAE5DEE8 for <idr@merit.edu>; Thu, 5 Dec 2002 12:15:07 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB5HF5b60059 for idr@merit.edu; Thu, 5 Dec 2002 12:15:05 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB5HEtC60028 for <idr@merit.edu>; Thu, 5 Dec 2002 12:14:55 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message Subject: Response to FSM input MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Date: Thu, 5 Dec 2002 12:14:55 -0500 X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AB09@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: need your response to the attached messages Thread-Index: AcKY3/TzEehR67FBQvykks4YJlFf/wDl3aWQ From: "Susan Hares" <shares@nexthop.com> To: "Yakov Rekhter" <yakov@juniper.net>, "Jeffrey Haas" <jhaas@nexthop.com>, <siva@ctd.haltech.com> Cc: <idr@merit.edu> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id MAA08898 Siva: Thanks for the comments. It helps to catch all the nits in the specification. I've sent response to your comments 1-10 in this email. I'll send the rest of the replies in groups of 10. Here's a summary of the Comment 1 - no consensus Comment 2a - out of date (No IdleHold Timer exists) Comment 2b - no consensus, can't remove due to OpenSent state action. Comment 3 - no text proposed, do we want all options specified in 8.1 (7 options) Please review and comment on text Comment 4 - agree Comment 5 - no text proposed, no consensus Comment 6 - no consensus on the text Comment 7 - see comment 3, no consensus Comment 8 - question, no text Comment 9 - agree, I supplied text Comment 10 - agree to remove text in option, different text for definition. Sue Hares ========== comment 1: 1) removal of peer oscillation damping The reason we should not remove the peer oscillation damping from the state machine: 1) Deployed implementations support peer oscillation damping 2) Hooks for the additions in the FSM cannot be added later. These hooks are optional and do not need to be implemented. 2) Idle HoldTimer in draft-18 Perhaps this is an old comment. I don't find IdleHoldTimer in the messages. 3) HoldTime and HoldTimer The Hold Time is specified on page 13 of the draft-18 specification as to the value and size. The text in OpenSent requires both a HoldTime and HoldTimer value. =========== OpenSent: In this state BGP waits for an Open Message from its peer. When an OPEN message is received, all fields are checked for correctness. If there are no errors in the OPEN message [Event 18] the local system: - resets the Open Delay timer to zero, - reset BGP Connect Timer to zero, - sends a KEEPALIVE message and - sets a KeepAlive timer (via the text below) - sets the hold timer according to the negotiated value (see Section 4.2), and - changes its state to OpenConfirm. If the negotiated hold time value is zero, then the Hold and KeepAlive timers are not started. If the value of the Autonomous System field is the same as the local Autonomous System number, then the connection is an "internal" connection; otherwise, it is an "external" connection. (This will impact UPDATE processing as described below.) =================== The earlier comments on the list indicated we need to collect what timers or Timer values were utilized in the FSM. I believe Alex made this comment last January. (the list archives will have this information.) So, I don't see any way to Comment 3: optional parameters specified Please suggest text. We have the following optional parameters: 1) Delay Open 2) Accept unconfigured peers 3) automatic start [Event 3] 4) passive TCP establishment (optional) 4 + 5 = Event 5 5) BGP_stop_flap (BGP Peer oscillation) [Event 6] 6)automatic stop [Event 7] 7)Collision detect in Established State If you want some of the optional parameters specified, let's catch all. Please suggest text and location. It could go in 8.1. -------------------------------- Comment 4 Old Text: Event 1: Manual Start Definition: Administrator manually starts peer connection. ... Event 2: Manual Stop Definition: Local system administrator manually stops the peer connection. New Text: Event1: Manual start Definition: Local system administrator manually starts the bgp connection. ... Event 2: Manual stop: Definition: Local system administrator manually stops the bgp connection. agree? --------------- Comment 5: 5) Event3, Event5, Event6 and Event7 5a) ========= Event 3/6/7 - cisco indicated the automatic stop, see the NANOG archives for a discussion of the Bad Route http://www.merit.edu/mail.archives/nanog/2001-06/msg00805.html The "event" sequences are new with this version of the FSM to identify particular events by number. 5b ====== Can we give examples of these events, as implemneted by curent implementations ? This will be of help to new implementaions. An example of Auto Stop is already present in the Update event handling section, and we can move that to this place as well. I believe that examples are outside the scope of this specification. If you have text clarifying the event descriptions, please send it in. ======================================================= 6) Event 4,5 Definition > Definition: Administrator manually start the peer > Connection, but has the passive flag=20 > enabled. The passive flag indicates=20 > that the peer will listen prior to=20 > establishing the connection. > Definition: Local system automatically starts the > BGP session with the passive flag=20 > enabled. The passive flag indicates=20 > that the peer will listen prior to=20 > establishing a connection. >My understanding is this is used to indicate a peer configuration >where we start the state machine, and wait for an incoming connection >request from the configured peer. Can we re-phrase this section to >indicate this. >How about something like: > >% The passive flag indicates that the state machine will wait for >% specified peer to initiate a connection with the local system. If >% this does not happen within a specific time (hold time), the local >% system will then also attempt to initiate connection with the >% specified peer. The text in 8.2.1.1 indicates the definition of the passive flag. 6a) ========== My understanding of your text is that you want to replace in both sets of text: "The passive flag indicates the peer will listen prior to establishing a connection". with: "The passive flag indicates that the state machine will wait for the specified peer to initiate a connection with a local system. The problem with this sentence is that in the "unconfigured" case the phrase "specified" peer is confusing. I think the original text is clearer. 6b) ========== If this does not happen within a specific time (hold time), the local system will then also attempt to initiate (a) connection with the specified peer. My comments: Again, the "specified peer" term is confusing. Also, the 2nd half of the statement mixes the actions of the state machine with the events. I believe this muddies the text instead of clarifying it. ================= Comment 7 - Event 6 - (not specified, please confirm) > Won't a variant of this with bgp_stop_flap option set be requried ? > We can also achieve the same by using the bgp_stop-Flap option as a > flag that is provided as an input to the state machine. see comment number 3. bgp_stop_flap is an option. This state machine description has many events to describe the mixture of options and actions. I believe it has a better clarity than testing lots of flags within a state. A proof in case, is that many people are making comments about specific transitions. ============ Comment 8 - 8) Event 5 Definition > Definition: Local system automatically starts the > BGP session with the passive flag=20 > enabled. The passive flag indicates=20 > that the peer will listen prior to=20 > establishing a connection. > Is it correct that this event will occur only when we want to restart > a connection (after it had been terminated due to some reason beside > administrative action) that we had accepted from an unconfigured peer ? The automatic start function is an implementation specific mechanism. This text does not seek to restrict it in any fashion. ================ 9) Timer Events Definition > Event8: idle Hold timer expires=20 > > Definition: Idle Hold timer expires. The Idle=20 > > Event9: connect retry timer expires > > Definition: An event triggered by the expiration of=20 > Event10: hold time expires > > Definition: An event generated when the HoldTimer=20 We can use any one of these terms (trigerred by/generated when) for all timer event definitions to make it look more consistent. agree: Will use "generate" Event 8: An event generated when the Idle Hold timer expires. Event 9: An event generated when the ConnectRetry timer expires. Event 10: An event generated when the Hold timer expires. Event 11: An event generated when the Keepalive timer expires Event 12: An event generated when the Delay BGP Open timer expires. Please confirm that this change to the definitions is agreeable to you. ================ 10) Event 8 Definition > Event8: idle Hold timer expires=20 > > Definition: Idle Hold timer expires. The Idle=20 > Hold Timer is only used when persistent =20 > BGP oscillation damping functions are=20 > enabled. =20 > > Status: Optional. Used when persistent > BGP peer oscillation damping functions > are enabled.=20 The status repeats the senetnce in the definition. Can we remove the sentence from the Status, or replace with a different definition. Do you think the defenition below would fit ? % An event trigerred by the expiry of the Idle Hold timer, indicating % that the session has completed waiting for a back-off period to % prevent flap, and may now proceed to take any further systemm % initiated action for this session. text change suggestion for 1st sentence that I'll agree to. An event trigerred by the expiry of the Idle Hold timer, indicating that the session has completed waiting for a back-off period to prevent bgp peer oscillation. Please confirm that you will agree to this. ================= End of comments on 1st 10 issues. Sue ------- Message 4 Date: Fri, 29 Nov 2002 19:46:06 +0530 From: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com> To: skh@nexthop.com, Yakov Rekhter <yakov@juniper.net> cc: skh@ndzh.com Subject: Comments on draft 18 This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. - ------_=_NextPart_000_01C297B1.DB0F1700 Content-Type: text/plain; charset="iso-8859-1" Hello, Attached are some comments that I have on draft-ietf-idr-bgp4-18. Please have a look at these and tell me if the same can be raised to the mailing list as well. Thanks, Siva (siva@ctd.hcltech.com) - ------_=_NextPart_000_01C297B1.DB0F1700 Content-Type: text/plain; name="BGP_Issues.txt" Content-Transfer-Encoding: quoted-printable Content-Disposition: attachment; filename="BGP_Issues.txt" All quoted sections of the original document have been prefixed by a '> = ' All lines of proposed text have been prefixed by a '% ' 1) Performing of peer oscillation damping: Since this feature is going to be added in a new draft, and its addition will change the operation of the state machine, can we remove all mention of it in the state machine ? As part of this removal, can we also remove the IdleHold Timer from the FSM since it is not useful in the absence of peer oscillation damping ? The draft that describes this procedure can then describe the change in the state machine required to do this. 2) Session Attributes: > Session Attributes required for each connection are;=20 > 1) state > 2) Connect Retry timer > 3) Hold timer > 4) Hold time > 5) Keepalive timer Why have a Hold Timer and a Hold Time ? Can we replace this with just Hold timer ? Can we also add the following session attributes: a) DelayBgpOpenTimer b) IdleHold Timer (in case we choose not to remove this from the bas = FSM) Can we also add the following flag to the session attributes: a) DelayOpen Flag 3) All sessions attributes: There are some attributes that are common to all sessions of the = system. Can we call these out as well: * Accept Connections from unconfigured peers (Enabled/Disabled) * Peer Oscillation Dampening (Enabled/Disabled) (In case we choose not to remove it from base spec) 4) Description for Event1/Event2 > Event1: Manual start=20 > > Definition: Administrator manually starts peer > connection > Event2: Manual stop > Definition: Local system administrator manually=20 > stops the peer connection. We can make these look consistent by using the Term Administrator or Local System administrator in both cases. 5) Event3, Event5, Event6 and Event7 Can we give examples of these events, as implemneted by curent implementations ? This will be of help to new implementaions. An example of Auto Stop is already present in the Update event handling section, and we can move that to this place as well. 6) Event 4,5 Definition > Definition: Administrator manually start the peer > Connection, but has the passive flag=20 > enabled. The passive flag indicates=20 > that the peer will listen prior to=20 > establishing the connection. > Definition: Local system automatically starts the > BGP session with the passive flag=20 > enabled. The passive flag indicates=20 > that the peer will listen prior to=20 > establishing a connection. My understanding is this is used to indicate a peer configuration where we start the state machine, and wait for an incoming connection request from the configured peer. Can we re-phrase this section to indicate this. How about something like: % The passive flag indicates that the state machine will wait for % specified peer to initiate a connection with the local system. If % this does not happen within a specific time (hold time), the local % system will then also attempt to initiate connection with the % specified peer. 7) Event 4,5 Won't a variant of this with bgp_stop_flap option set be requried ? We can also achieve the same by using the bgp_stop-Flap option as a flag that is provided as an input to the state machine. 8) Event 5 Definition > Definition: Local system automatically starts the > BGP session with the passive flag=20 > enabled. The passive flag indicates=20 > that the peer will listen prior to=20 > establishing a connection. Is it correct that this event will occur only when we want to restart a connection (after it had been terminated due to some reason beside administrative action) that we had accepted from an unconfigured peer ? 9) Timer Events Definition > Event8: idle Hold timer expires=20 > > Definition: Idle Hold timer expires. The Idle=20 > > Event9: connect retry timer expires > > Definition: An event triggered by the expiration of=20 > Event10: hold time expires > > Definition: An event generated when the HoldTimer=20 We can use any one of these terms (trigerred by/generated when) for all timer event definitions to make it look more consistent. 10) Event 8 Definition > Event8: idle Hold timer expires=20 > > Definition: Idle Hold timer expires. The Idle=20 > Hold Timer is only used when persistent =20 > BGP oscillation damping functions are=20 > enabled. =20 > > Status: Optional. Used when persistent > BGP peer oscillation damping functions > are enabled.=20 The status repeats the senetnce in the definition. Can we remove the sentence from the Status, or replace with a different definition. Do you think the defenition below would fit ? % An event trigerred by the expiry of the Idle Hold timer, indicating % that the session has completed waiting for a back-off period to % prevent flap, and may now proceed to take any further systemm % initiated action for this session. 11) Event 10 (Hold timer) We use the hold timer for two purposes * Waiting for an open message (with a default value of 240 seconds) * Waiting for Keepalives (with a default value of 90 seconds) Can we use two different timers (or at least call them two different timer events) ? 12) Event 12 > Event12: DelayBGP Open timer expires [changed] >=09 > Definition: A timer that delays sending of the BGP > Open message for n seconds after the > Transport connection has been completed.=20 The Definition seems to better define what the timer is, rather than = what its expiry means. Can we change it to reflect this ? 13) Event 13, 14 definition > Event13: Transport Connection Indication & valid remote peer[changed] > > Definition: Event indicating that TCP connection > request with a valid source IP address and TCP=20 > port, and valid destination IP address=20 > and TCP Port. The definition of=20 > invalid Source, and invalid Destination=20 > IP address is left to the implementation. a) How about changing this slighty to: % Definition: Event indicating receiving a TCP connection % request with a valid source IP address and TCP=20 % port, and valid destination IP address=20 % and TCP Port. The definition of=20 b) Is it correct that Destination IP address validation is used when we do not wish to accept BGP connections on one of the local IP addresses (either because we wish to accept such addresses on only our loopback interfaces, or because the address to which we received the connection was not the optimal interface with which to speak to the sender) ? If so, can we add this to the description ? 14) Event 13,14 > Event13: Transport Connection Indication & valid remote peer[changed] > > Definition: Event indicating that transport > Request valid source IP address and TCP=20 > port, and valid destination IP address=20 > and TCP Port. The definition of=20 > invalid Source, and invalid Destination=20 > IP address is left to the implementation. a) I am unclear as to what validation of source and destination port = mean. The source port's value should be of no concern to connection = receipient system. The destination port cannot be invalid, since BGP would not receive the connection request in such a situation. b) I think it will be useful to state what we consider to be an invalid IP address. 15) Event 17 > Event17: TCP connection fails This event can occur even when the transport connection is closed by the other end. Since this does not reflect a 'failure ', can we change the event name to % Event17: TCP connection termination 16) Events related to BGP Messages Can we make the definition style across these events consistent ? 17) Event 19 Definition > Event19: BGPOpen with BGP Delay Open Timer running [changed] > > Definition: An event indicating that a valid Open=20 > Message has been successful=20 > established for a peer that is=20 > currently delaying the sending of an=20 > BGP Open message.=20 The definition appears a little fragmented. How about: % Definition: An event indicating that a valid OPEN % Message has been received for a peer % that has a successfully established % transport connection and is currently % delaying the seending of a BGP open % message 18) Event 22 Definition a)=20 > Event22: Open collision discard > > Definition: An event generated administratively=20 > when a connection Collision has been=20 > detected while processing an incoming =20 > Open message. This connection has been=20 Isn't this event 'automatically' generated, since it is a system generated event ?=20 b) > Event 22 is an administrative could=20 > occur if FSM is implemented as two=20 > The word event is missing. How about % Event 22 is an automatic event that % could occur if FSM is implemented as two=20 19) FSM Description: ConnectRetry Count The Connect Retry count is updated by the FSM but never used. In the absence of peer oscillation damping, will this be used to stop connection establishment attempts after a certain maximum number ? Can we include the use of this counter in some place ? 20) Handling Event 7 (Auto Stop) to Idle State processing: The handling of Event 7 is missing from the Idle State processing. Can we add this ? How about replacing > An manual stop event (Event2) is ignored in the Idle state. with % Manual stop (Event 2) and Auto stop (Event 7) events are ignored % in the Idle state 21) Clearing of Connection Retry Timer: There are a few sections where the FSM draft states that the Connection Retry timer needs to be reset, wheras the conenct retry timer had been cleared prior to entering that state. We can remove these instructions to clear the connect retry timer. List of places where the connect retry timer need not be cleared a) Handling of Event 19 in the Connect State b) Handling of Events 12 in the Active State c) All cases where it is refered to in the OpenSent, OpenConfirm and Established states 22) Handling of Event 14 in the Connect state > If the transport connection receives an indication > that is invalid or unconfigured. [Event 14]: > - the TCP connection is rejected. I don't understand how we would get this event while in this state. 23) Handling events 20, 21 in the Connect State and Active State We need to consider the case where we receive events 20 (message header error) and 21 (Open message error) when the delay timer is = running. Since the connection has been established at this point, we need to = send a Notification message and then terminate the connection. 24) Handling the default events in the Connect state The BGP Delay OpenTimers needs to be cleared if it is running. How about adding this: % - If the ConnectRetry Timer is running % - Clear the Connect Retry timer % - Otherwise % - Clear the BGP Delay Open Timer 25) Handling Event 23 (Notify with version mismatch error) in Connect and OpenSent This is currently being handled in the default event processing = section. However, we do not need to go through the peer oscillation damping process in this case. Can we change the wordings to reflect this, or move this out of peer oscillation damping processing ? 26) Handling Event 17 in The connect state > If the transport connection fails (timeout or transport > disconnect) [Event17], the local system: > - changes its state to Active.=20 If the transport connection fails when the BGP Open Delay timer is running, should we still be going into the Active state ? 27) Handling Event 17 in Active state We should now move into Idle state. Can we add % - Goes to Idle state 28) Handling Event 19 in the Active state > - Set the Hold timer to a large value (4 minutes), Since OPEN messages have been exchanged, can we change this to % - If the negotiated Hold time is not 0, set the Hold time to % - the negotiated value 29) Handling Event 2 in the Active State > A manual stop event[Event2], the local system: > - Sends a notification with a Cease, > - drops the Transport connection,=20 These two actions are possible only if a transport connection had = already been established. How about changing the text to % - If a transport connection had been successfully established % - Send a Notification with a Cease % - Drop the Transport Connection 30) Default event handling in Active State To ensure consistencey with E2 handling, can we add % - If any BGP Routes exist, delete the routes 31) Clearing Hold timer in OpenSent, OpenConfirm and Established State In all event handling where we go to Idle state, we need to clear the Hold Timer as well. 32) Clearing Keepalive timer in OpenConfirm and Established State In all event handling where we go to Idle state, we need to clear the Keepalive Timer as well. 33) Handling Event 18 in the OpenSent state Why do we start the Keepalive timer at this stage ? Isn't it sufficient = to do so when we move into Established state ? 34) Established State Some event handling in the Established state do not set the MIB Reason = when handling an event that causes an error. Can we add this ? General Comments 35) For the events whose status is optional, can we state the impact of = not supporting them (in terms of any interoperability issues). I = understand that most of the optional events will not have such an impact; but a = clarification statement for the optional events would benefit new implementors. 36) Is delaying the sending of an OPEN message a standrad industry = practice ? Also, in the FSM, this has been handled by practically implementing = a sub-state each, within the CONNECT and ACTIVE states. Won't the FSM look more simple if we just had a new DelayOpen state that we could move into ? - ------_=_NextPart_000_01C297B1.DB0F1700-- -- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id JAA28570 for <idr-archive@nic.merit.edu>; Thu, 5 Dec 2002 09:20:26 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 5E907912CC; Thu, 5 Dec 2002 09:20:01 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 25D43912CD; Thu, 5 Dec 2002 09:20:01 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id C15C5912CC for <idr@trapdoor.merit.edu>; Thu, 5 Dec 2002 09:19:59 -0500 (EST) Received: by segue.merit.edu (Postfix) id A65515DF2E; Thu, 5 Dec 2002 09:19:59 -0500 (EST) Delivered-To: idr@merit.edu Received: from ietf.org (odin.ietf.org [132.151.1.176]) by segue.merit.edu (Postfix) with ESMTP id 078425DF16 for <idr@merit.edu>; Thu, 5 Dec 2002 09:19:59 -0500 (EST) Received: from CNRI.Reston.VA.US (localhost [127.0.0.1]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id JAA12304; Thu, 5 Dec 2002 09:17:07 -0500 (EST) Message-Id: <200212051417.JAA12304@ietf.org> Mime-Version: 1.0 Content-Type: Multipart/Mixed; Boundary="NextPart" To: IETF-Announce: ; Cc: idr@merit.edu From: Internet-Drafts@ietf.org Reply-To: Internet-Drafts@ietf.org Subject: I-D ACTION:draft-ietf-idr-aspath-orf-04.txt Date: Thu, 05 Dec 2002 09:17:05 -0500 Sender: owner-idr@merit.edu Precedence: bulk --NextPart A New Internet-Draft is available from the on-line Internet-Drafts directories. This draft is a work item of the Inter-Domain Routing Working Group of the IETF. Title : Aspath Based Outbound Route Filter for BGP-4 Author(s) : K. Patel, S. Hares Filename : draft-ietf-idr-aspath-orf-04.txt Pages : 0 Date : 2002-12-4 This document defines a new Outbound Router Filter type for BGP, termed 'Aspath Outbound Route Filter', that can be used to perform aspath based route filtering. This ORF-type supports aspath based route filtering as well as regular expression based matching, for address groups. A URL for this Internet-Draft is: http://www.ietf.org/internet-drafts/draft-ietf-idr-aspath-orf-04.txt To remove yourself from the IETF Announcement list, send a message to ietf-announce-request with the word unsubscribe in the body of the message. Internet-Drafts are also available by anonymous FTP. Login with the username "anonymous" and a password of your e-mail address. After logging in, type "cd internet-drafts" and then "get draft-ietf-idr-aspath-orf-04.txt". A list of Internet-Drafts directories can be found in http://www.ietf.org/shadow.html or ftp://ftp.ietf.org/ietf/1shadow-sites.txt Internet-Drafts can also be obtained by e-mail. Send a message to: mailserv@ietf.org. In the body type: "FILE /internet-drafts/draft-ietf-idr-aspath-orf-04.txt". NOTE: The mail server at ietf.org can return the document in MIME-encoded form by using the "mpack" utility. To use this feature, insert the command "ENCODING mime" before the "FILE" command. To decode the response(s), you will need "munpack" or a MIME-compliant mail reader. Different MIME-compliant mail readers exhibit different behavior, especially when dealing with "multipart" MIME messages (i.e. documents which have been split up into multiple messages), so check your local documentation on how to manipulate these messages. Below is the data which will enable a MIME compliant mail reader implementation to automatically retrieve the ASCII version of the Internet-Draft. --NextPart Content-Type: Multipart/Alternative; Boundary="OtherAccess" --OtherAccess Content-Type: Message/External-body; access-type="mail-server"; server="mailserv@ietf.org" Content-Type: text/plain Content-ID: <2002-12-4093302.I-D@ietf.org> ENCODING mime FILE /internet-drafts/draft-ietf-idr-aspath-orf-04.txt --OtherAccess Content-Type: Message/External-body; name="draft-ietf-idr-aspath-orf-04.txt"; site="ftp.ietf.org"; access-type="anon-ftp"; directory="internet-drafts" Content-Type: text/plain Content-ID: <2002-12-4093302.I-D@ietf.org> --OtherAccess-- --NextPart-- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id VAA18645 for <idr-archive@nic.merit.edu>; Wed, 4 Dec 2002 21:25:15 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 1A0CE912C3; Wed, 4 Dec 2002 21:24:49 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id DF9A7912C4; Wed, 4 Dec 2002 21:24:48 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 5644E912C3 for <idr@trapdoor.merit.edu>; Wed, 4 Dec 2002 21:24:47 -0500 (EST) Received: by segue.merit.edu (Postfix) id 338ED5DEB1; Wed, 4 Dec 2002 21:24:47 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 867815DE99 for <idr@merit.edu>; Wed, 4 Dec 2002 21:24:46 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB52Oj244159 for idr@merit.edu; Wed, 4 Dec 2002 21:24:45 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB52OdC44145 for <idr@merit.edu>; Wed, 4 Dec 2002 21:24:39 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message Subject: fsm missing next state - Event 14 in Connect state : Implementers - review needed MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Date: Wed, 4 Dec 2002 21:24:39 -0500 X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AB00@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: fsm missing next state - Event 14 in Connect state : Implementers - review needed Thread-Index: AcKcBXY6q+Y3GYOeRhOVBwM78XFN4Q== From: "Susan Hares" <shares@nexthop.com> To: <idr@merit.edu> Cc: <nwnetworks@dial.pipex.com>, <nwnetworks@dial.pipex.com>, "Jeffrey Haas" <jhaas@nexthop.com> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id VAA18645 Calling all implementers, please comment on this proposed change. Connect State - Event 14 [no consensus] Current Text: If the TCP connection receives an indication that that is invalid or unconfigured [Event 14], - the TCP connection is rejected. At the very least this section needs more "word smithing", so I'd like to change it for more clarity at least. I'm not sure this represents the implementations. What I'd like to do is query the implementations to see what they do if they receive a valid TCP connection with an invalid or unconfigured peer. Two options: Alternative 1: Count it as a valid response New Text: If a TCP connection is received that has an invalid format, or an unconfigured host [Event 14], the local system: - rejects the TCP connection, - increments the connect retry counter, - performs bgp peer oscillation checks. If bgp peer oscillation checks allow for a new connection, the bgp peer - restarts the Connect retry timer with configured value, and - enters the Active state. FSM table: Idle Connect Active Open-Sent Open-Confirm Establish Event-14 ======================================================= Nxt state Idle | Active|Active|Open-Sent|Open-Confirm|Establish| ======================================================= action V | Y2 | L | Ignore | Ignore | Ignore | ======================================================= Alternative 2: Reject the connection and see if valid or configured one appears within the connect timer window. New Text: If a TCP connection is received that has an invalid format, or an unconfigured host [Event 14], the local system: - rejects the TCP connection, - and stays in the Connect state. FSM table: Idle Connect Active Open-Sent Open-Confirm Establish Event-14 ======================================================= Nxt state Idle |Connect|Active|Open-Sent|Open-Confirm|Establish| ======================================================= action V | L | L | Ignore | Ignore | Ignore | ======================================================= Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id VAA18284 for <idr-archive@nic.merit.edu>; Wed, 4 Dec 2002 21:19:04 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 40EBF912C2; Wed, 4 Dec 2002 21:18:46 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 1205F912C3; Wed, 4 Dec 2002 21:18:45 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 6A5C2912C2 for <idr@trapdoor.merit.edu>; Wed, 4 Dec 2002 21:18:44 -0500 (EST) Received: by segue.merit.edu (Postfix) id 52A1A5E100; Wed, 4 Dec 2002 21:18:44 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 9B9685E0FB for <idr@merit.edu>; Wed, 4 Dec 2002 21:18:43 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB52IgE44042 for idr@merit.edu; Wed, 4 Dec 2002 21:18:42 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB52IUC44007 for <idr@merit.edu>; Wed, 4 Dec 2002 21:18:30 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message Subject: fsm missing next state - Events 13 - 17 and the TCP connection MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Date: Wed, 4 Dec 2002 21:18:30 -0500 X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AAFF@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: fsm missing next state - Events 13 - 17 and the TCP connection Thread-Index: AcKcBJocOcelZzGqSwmzlHSQ0QCkoA== From: "Susan Hares" <shares@nexthop.com> To: <idr@merit.edu> Cc: <nwnetworks@dial.pipex.com>, "Jeffrey Haas" <jhaas@nexthop.com>, <zinin@psg.com>, <yakov@juniper.net> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id VAA18284 Implementers of the BGP FSM we need your help! Tom Petch has raised a concern about events 13-17 on the mail list in thread: RE: bgp18 WG Last Call fsm missing next state I welcome any comments public (or to me or Yakov privately) on this topic. I will utilize private comments to answer some of the questions below. ============================ These events are described in section 8.1.3. In our discussion in January through May of 2002, many implementers mapped their implementation onto the following TCP events list in 8.1.3. Events 13 - 17 Event 13 - TCP connection indication & valid remote peer Event 14 - TCP connection indication with invalid source or destination Event 15 - TCP connection request sent (by this peer) received an Acknowledgement [ local system sent a TCP SYN, Received a TCP SYN, ACK pair back, and Sent a TCP ACK] Event 16 - TCP connection confirmed [local system received a TCP SYN, sent a TCP SYN, ACK back, and received a TCP ACK] Event 17 - TCP connections Should we have all of these states? Which implementations support all of these Events? The FSM State machine for this group is Idle Connect Active Open-Sent Open-Confirm Establish Event-13: [TCP connection request from valid remote peer] ======================================================= Nxt state Idle |Connect|Active|Open-Sent|Open-Confirm|Establish | ======================================================== action V | *1 | *2 |Track 2nd | Track 2nd | Track 2nd | | | |connection| connection | connection ========================================================= V - Process FSM Error. * 1 We sent TCP SYN. Is this TCP connection in response or a new connection to port 179? The FSM state machine in the hares-bgp4-statemt-02 draft uses this event to track the TCP connection messages. The sub-state machine uses flags 1) Await TCP SYN 2) Sent TCP SYN 3) Await TCP ACK 4) Null - no flags These state is being used to track the TCP connection state. recommendation: Make this Event optional. Open questions: Do we believe the Two TCP SYN cross for the same connection? Does this TCP related portion of the state machine handle this? * 2 - local system received a TCP SYN from the remote port. All the same comments from *1 apply. Event 14 TCP REQ request received from an invalid peer ======================================================= Current Idle Connect Active Open-Sent Open-Confirm Establish ======================================================= Nxt state Idle |Connect|Active|Open-Sent|Open-Confirm|Establish| ======================================================= action V | L | L | Ignore | Ignore | Ignore | ======================================================= Event 15 - TCP connection request sent received an ACK ======================================================= Current Idle Connect Active Open-Sent Open-Confirm Establish ======================================================= Next state Idle |Connect|Active |Open-Sent|Open-Confirm|Establish| |or Open|or Open | | | | | Sent | Sent | | | | ========================================================= action V | *3 | *4 | Ignore | Ignore | Ignore | ========================================================= *3/*4 - The actions taken in *3/*4 differ only in the next state if the Delay Open flag is set. - If Delay Open flag is set, the local system will: do (ZZ): - Clear the connection retry timer, and - Set Open Delay timer to initial value, and - [*3]stay in the Connect state [*4]stay in the Active state - If Delay Open flag is not set, the local system will: do (H) - clears the connect retry timer, - completes the BGP initialization, - clears the Open Delay timer (sets to zero), - send an Open message to its peer, - sets hold timer to a large value, and - change the state to Open Sent. Event 16 - TCP connection confirmed (Received TCP Req, Sent TCP SYN, ACK), Current Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Connect|Active |Open-Sent|Open-Confirm|Establish| |or Open|or Open | | | | | Sent | Sent | | | | ========================================================= action V | *3 | *4 | Ignore | Ignore | Ignore | ======================================================== Event 17 - TCP Connection Fails ======================================================== Current Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Nxt state Idle |Active |Idle |Active | Idle |Idle | ========================================================= action V | G | Y2 | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= G: the local system will: - restart the connect retry timer (with initial value), - continue to listen for a connection that may be initiated by remote peer, and - change its state to active. Y2 - is: The local system: 1) increments the ConnectRetryCnt (connect retry count) by 1, 2) Releases all BGP resources associated with this connection, and 3) performs the BGP peer oscillation damping process. If the damping process allows for a new connection, the local system - restarts the connect retry timer (with initial value), and - goes to Idle If the damping process does not allow for a new connection, the local system - releases the BGP resources, and goes to - Idle. Please note: the proposal for event 17, has Connect and Active doing the same thing. Please comment on how your state machine works. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id VAA18229 for <idr-archive@nic.merit.edu>; Wed, 4 Dec 2002 21:18:23 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id DDFA291205; Wed, 4 Dec 2002 21:18:03 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id A5868912C2; Wed, 4 Dec 2002 21:18:03 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 4D77991205 for <idr@trapdoor.merit.edu>; Wed, 4 Dec 2002 21:17:54 -0500 (EST) Received: by segue.merit.edu (Postfix) id 283C95E103; Wed, 4 Dec 2002 21:17:54 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 9A3C45E102 for <idr@merit.edu>; Wed, 4 Dec 2002 21:17:53 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB52Hq343955 for idr@merit.edu; Wed, 4 Dec 2002 21:17:52 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB52HkC43942 for <idr@merit.edu>; Wed, 4 Dec 2002 21:17:46 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message Subject: RE: bgp18 WG Last Call fsm missing next state MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Date: Wed, 4 Dec 2002 21:17:46 -0500 X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AAFE@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: bgp18 WG Last Call fsm missing next state Thread-Index: AcKaHxSLTG7nC0wCT3C+gannXhGWHwBtZ4Kg From: "Susan Hares" <shares@nexthop.com> To: "Tom Petch" <nwnetworks@dial.pipex.com>, "Jeffrey Haas" <jhaas@nexthop.com> Cc: "idr" <idr@merit.edu> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id VAA18229 Tom and Jeff: Thanks for the great review of the state machine. Please confirm I've got my 7 changes numbered right. Here's where we have consensus: change 1 - consensus, editorial state: Connect Events: 15 and 16 change 2 - no consensus I've given two alternatives. We need to get implementers to pick 1. thread: State: Connect Event: 14 change 3 - consensus, editorial state: Active: Events: 15 and 16 change 4 and 5: State: Active Events: 13 and 14 On changes 4 and 5 - moved to a different thread thread titled: fsm missing next state - Events 13 - 17 and the TCP connection on change 6 - no consensus State: Active Event: 17 Review my explanation and the new text, and see if we have consensus. on change 7: no consensus. State: Open Confirm Event: 18 Review my explanation and see if we have consensus. Sue Hares (skh@nexthop.com) ============= Warning - put the screen on wide mode for the following text: ------------------- Here's my response on the 7 changes 1) Connect State - Event 15 or Event 16 [consensus, editorial] note: The delay retry timer is utilized instead of the connect retry timer for the next two changes. Previous text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the "Delay Open Flag". If the delay open flag is set, the local system: - clears the connect retry timer, - sets the BGP open delay timer to initial value If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clear BGP Open Delay timer (set to zero), - completes the BGP initialization, - send an Open message to its peer, - sets hold timer to a large value, and - change the state to Open Sent. New text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the Delay Open flag prior to processing: If the Delay Open flag is set, the local system: - clears the connect retry timer, - sets the BGP open delay timer to initial value, and - stays in the Connect state. If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clears the BGP Delay timer (sets to zero), - completes the BGP initialization, - sends an Open message to its peer, - sets the hold timer to a large value, and - changes the state to Open Sent. 2)Connect State - Event 14 [no consensus] Current Text: If the TCP connection receives an indication that that is invalid or unconfigured [Event 14], - the TCP connection is rejected. At the very least this section needs more "word smithing", so I'd like to change it for more clarity at least. I'm not sure this represents the implementations. What I'd like to do is query the implementations to see what they do if they receive a valid TCP connection with an invalid or unconfigured peer. Two options: Alternative 1: Count it as a valid response New Text: If a TCP connection is received that has an invalid format, or an unconfigured host [Event 14], the local system: - rejects the TCP connection, - increments the connect retry counter, - performs bgp peer oscillation checks. If bgp peer oscillation checks allow for a new connection, the bgp peer - restarts the Connect retry timer with configured value, and - enters the Active state. FSM table: Idle Connect Active Open-Sent Open-Confirm Establish Event-14 ======================================================= Next state Idle | Active|Active|Open-Sent|Open-Confirm|Establish| ======================================================= action V | Y2 | L | Ignore | Ignore | Ignore | ======================================================= Alternative 2: Reject the connection and see if valid or configured one appears within the connect timer window. New Text: If a TCP connection is received that has an invalid format, or an unconfigured host [Event 14], the local system: - rejects the TCP connection, - and stays in the Connect state. FSM table: Idle Connect Active Open-Sent Open-Confirm Establish Event-14 ======================================================= Nxt state Idle |Connect|Active|Open-Sent|Open-Confirm|Establish| ======================================================= action V | L | L | Ignore | Ignore | Ignore | ======================================================= 3) Active State - Event 15 or Event 16 [consensus, editorial] Previous text: A TCP connection succeeds [Event 15 or Event 16], the local system: process the TCP connection flags - If the BGP delay open flag is set: - clears the connect retry timer [through the following text: - and changes its state to Open Sent. New text: If the TCP connection succeeds [Event 15 or Event 16], local system checks the "Delay Open Flag" prior to processing: If the delay open flag is set, the local system: - clears the connect retry timer, - sets the BGP open delay timer to initial value, and - stays in the Active state. If the Delay Open flag is not set, the local system: - clears the connect retry timer, - clears the BGP Delay timer (sets to zero), - completes the BGP initialization, - sends an Open message to its peer, - sets the hold timer to a large value, and - changes the state to Open Sent. 4) Active State, Event 13 [no consensus] 5) Active State, Event 14 [no consensus] The problem with this state is it is difficult to exactly specify without discussing the TCP Messages that FSM document covers. I'll query if the implementors require all of events 13-17 as mandatory. We'll 6) Active State, Event 17 [no consensus] comment: In the Active state, we may already have a connection and be awaiting the Open Delay timer. The TCP disconnect or timeout could occur in this state due to the "Open Delay Timer". If the TCP Disconnect is ignored, we could have some peer oscillation. If the we wait, then the connection retry timer needs to be kept running. The text below allows this timer. The real question is what is the status of the current implementations. I agree, the Active state and the connect state should match. Old Text: If the TCP connection fails (timeout or disconnection) [Event 17], the local system: - set TCP disconnect in the MIB reason code, - restart Connect retry timer (with initial value), - release all BGP resources, - Acknowledge the Drop of the TCP connection if TCP disconnect (FIN ACK), - Increment ConnectRetryCnt (connect retry count) by 1, and - performs the BGP peer oscillation damping process. Applicable FSM State table: FSM table old: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Active |Idle |Active | Idle |Idle | | | | | | | ========================================================= action V | Y2 | G | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= Alternative 1: FSM table new: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Active |Active |Active | Idle |Idle | | | | | | | ========================================================= action V | G | G | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= G: The local system: - restarts the connect retry timer (at intial value), - continues to listen for a connection that may be initiated by the remote peer, and - sets its next state to Active. New Text: (for Connect and Active state) If the TCP connection fails (timeout or disconnect) [Event 17], the local system: - restarts the connect retry timer, - continues to listen for a connection that may be initiated by the remote BGP peer, and - changes it state to Active. Alternative 2: FSM table new: Event 17 current: Idle Connect Active Open-Sent Open-Confirm Establish ========================================================= Next state Idle |Idle |Idle |Active | Idle |Idle | | | | | | | ========================================================= action V | Y2 | Y2 | Ignore| Track 2nd | Track 2nd | | | | | connection | connection| ========================================================= Next Text: If the location system receives a TCP connection failed [Event 17], the local system will: - increment the ConnectRetrycnt (connect retry count) by 1, - release all BGP resources associated with this connection, - perform BGP peer oscillation (if configured), and - go to Idle Y2 - is: The local system: 1) increments the ConnectRetryCnt (connect retry count) by 1, 2) releases all BGP resources associated with this connection, and 3) performs the BGP peer oscillation damping process if the damping process allows for a new connection, the local system - restarts the connect retry timer (with initial value, and - goes to Idle If the damping process does not allow for a new connection, the local system - set the flags to damp the creation of a new bgp connection until a manual start occurs, and - goes to Idle. 7) Open Confirm state, Event 18 Here's my proposed text. Please let me know what you think. I think this is an editorial change. Old text: If the open message is valid, the collision detect function is processed per section 6.8. If this connection is to be dropped due to call collision, the local system: - sends a Notification with a Cease - resets the Connect timer (to zero), - releases all BGP resources, - Drop the TCP connection (sends a TCP FIN), - increments the ConnectRetryCnt by 1 (connect retry count), and - performs an BGP peer oscillation damping process. New text: If the open message is valid, the BGP peer connection is check to detect a collision per section 6.8. If this connection is to be dropped due to call collision, the local system: - sends a Notification with a Cease - resets the Connect timer (to zero), - releases all BGP resources, - Drop the TCP connection (sends a TCP FIN), - increments the ConnectRetryCnt by 1 (connect retry count), and - performs an BGP peer oscillation damping processing, and - enters the Idle State. notes: Collision detect impacts Open Sent, Open Confirm, and Established states. Applicable FSM state >> Open Confirm State: >> >> If the Open messages is valid [Event 18], the collision >> detect function is processed per section 6.8. If this >> connection is to be dropped due to call collision, the >> local system: >>** enters what state > > This instance of the FSM should be discarded since it implies there is > another FSM instance running that will proceed to Established? -----Original Message----- From: Tom Petch [mailto:nwnetworks@dial.pipex.com] Sent: Monday, December 02, 2002 11:07 AM To: Jeffrey Haas Cc: idr Subject: Re: bgp18 WG Last Call fsm missing next state Jeff I appreciate your comments, agreeing with many. A minor supplementary; Open Confirm, valid open [event 18], decide to drop this connection and so the FSM but what about the actions of incrementing the ConnectRetryCnt (and optional peer oscillation damping)? Are we doing that in the other FSM? on a per router id or per IP address pair independent of FSM? And a slightly bigger issue in the Active state (which is the passive state because we have entered it from Idle with Passive TCP flag set Events 4/5). In which case, TCP Indication valid [event 13] or invalid [event 14] are what I would expect with event 14 keeping us in Active state; and if event 13 also keeps us in Active state (while sending a TCP SYN-ACK) then TCP Connection Confirmed [event 16] will ensue while TCP Connection Acknowledged [event 15] would be an FSM error. Except that you suggest that event 14 in the Connect state takes us to Active state which seems fine but for the fact that we will have sent a TCP SYN (because that is what the Connect state is), this invalid Indication may be a coincidence, in which case our SYN is about to receive a SYN-ACK back ie event 15 is valid after all. And I am then in agreement with you over events 15/16 in the Active (ie passive) state (delay open flag keeps us in the Active state). Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Jeffrey Haas <jhaas@nexthop.com> To: Tom Petch <nwnetworks@dial.pipex.com> Cc: idr <idr@merit.edu> Date: 30 November 2002 01:33 Subject: Re: bgp18 WG Last Call fsm missing next state >Tom, > >On Thu, Nov 28, 2002 at 02:53:56PM -0000, Tom Petch wrote: >> In seven places in the FSM, the actions listed for the occurrence of >> an event do not include any indication of what the next state should >> be, nor can I be sure I can guess what it should be. >> >> I believe the FSM should specify the new state in the cases listed >> below. >> Change 1: >> Connect State: >> >> If the TCP connection succeeds [Event 15 or >> Event 16], the local system checks the "Delay Open >> Flag". If the delay Open flag is set, the local system: >> **enters what state > >Remains in the connect state awaiting either an open message or >for the opendelay timer to expire. Agreed Change 2: > >> If the TCP connection receives an indication >> that is invalid or unconfigured. [Event 14]: >> **enters what state > >Enters the active state. >Connectretrytimer should be started. Agreed Change 3: >> Active State: >> >> A TCP connection succeeds [Event 15 or Event 16], the >> local system: process the TCP connection flags >> - If the BGP delay open flag is set: >> ** enters what state (I think this is an FSM error in TCP because it >> has not initiated a connection!) >In the active state, we are awaiting incoming tcp connections to be >completed. Here we should: > >remain in the active state >wait for an open message or the opendelay timer to expire. > Change 4: >> If the local system receives a valid TCP Indication >> [Event 13], the local system processes the TCP connection >> flags. >> ** enters what state > >I'm also confused by this event. Since we are not initiating a TCP >connection, but are awaiting one to complete, why would we get an >Indication here? FSM error? We would get an indication from the remote side that it had send us a TCP Syn. The real question is Change 5: >> If the local system receives a TCP indication >> that is invalid for this connection [Event 14]: >> ** enters what state > >Ditto. Change 6: >> >> If the local system receives a TCP connection >> failed [Event 17] (timeout or receives connection >> disconnect), the local system will: >> ** enters what state > >Ditto. Change 7: >> Open Confirm State: >> >> If the Open messages is valid [Event 18], the collision >> detect function is processed per section 6.8. If this >> connection is to be dropped due to call collision, the >> local system: >> ** enters what state > >This instance of the FSM should be discarded since it implies there is >another FSM instance running that will proceed to Established? > >> Tom Petch >> nwnetworks@dial.pipex.com > >-- >Jeff Haas >NextHop Technologies Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id OAA26115 for <idr-archive@nic.merit.edu>; Wed, 4 Dec 2002 14:50:40 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id A7A2A91232; Wed, 4 Dec 2002 14:50:12 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 6F5A3912BF; Wed, 4 Dec 2002 14:50:12 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id E7A7C91232 for <idr@trapdoor.merit.edu>; Wed, 4 Dec 2002 14:50:10 -0500 (EST) Received: by segue.merit.edu (Postfix) id C58245E0E0; Wed, 4 Dec 2002 14:50:10 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 324755DE1F for <idr@merit.edu>; Wed, 4 Dec 2002 14:50:10 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB4Jo7p34679 for idr@merit.edu; Wed, 4 Dec 2002 14:50:07 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB4Jo1C34666 for <idr@merit.edu>; Wed, 4 Dec 2002 14:50:01 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message Subject: RE: bgp18 WG Last Call fsm missing keepalive MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Date: Wed, 4 Dec 2002 14:50:01 -0500 Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AAFB@aa-exchange1.corp.nexthop.com> X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: bgp18 WG Last Call fsm missing keepalive Thread-Index: AcKbw0Li6ovNFLNkRbqWXgkZU9oIGQACM0UA From: "Susan Hares" <shares@nexthop.com> To: "Tom Petch" <nwnetworks@dial.pipex.com>, "idr" <idr@merit.edu> Cc: "Jeffrey Haas" <jhaas@nexthop.com> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id OAA26115 Tom: I agree about the putative OpenDelay, but I do not feel that represents the BGP implementations of today. We spent a long time Jan-Feb 2002 on adding any FSM States. My earlier note was too brief. Please answer questions 1) Agree, this can occur in the Active state as well as the Connect state. Will you accept the earlier text below to be inserted both places? Background: The state machine for Event 19 is: Idle Connect Active Open Open Estb Sent Confirm =============================================== Event 19 | | | | | | | next state |Idle | Open | Open | Open |Idle | Idle | | | confirm| confirm| Confirm| | | =============================================== action | V | N-2 | N-2 | N | E-1 | E | =============================================== Per the State Machine. Action v - FSM Error Action E - FSM Error, drop connection - etc, drop routes Action E-1 - FSM Error, drop connection (lots of Action N-2 (text below) Action N (text below, without sending Open) 2) Do you think that Event 19 is possible in the Open Sent state? Please answer this separately. Thanks for all the wonderful review of the State machine. I'm working my way through the comments. Sue -----Original Message----- From: Tom Petch [mailto:nwnetworks@dial.pipex.com] Sent: Wednesday, December 04, 2002 1:30 PM To: Susan Hares; idr Cc: Jeffrey Haas Subject: Re: bgp18 WG Last Call fsm missing keepalive Still not quite. This can occur in the Connect state as well as Active, when a TCP connection completes with delay open flag set, my putative OpenDelay state (sigh, point taken). And we should still go to OpenConfirm state at the bottom of the action list as is currently part of list H. Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Susan Hares <shares@nexthop.com> To: Tom Petch <nwnetworks@dial.pipex.com>; idr <idr@merit.edu> Cc: Jeffrey Haas <jhaas@nexthop.com> Date: 04 December 2002 18:07 Subject: RE: bgp18 WG Last Call fsm missing keepalive Tom: Just to clarify this comment: Event 19 = valid open with delay timer running **Connect or Active = 1) awaiting TCP connection, or 2) TCP connection completed and awaiting the TCP connection with delay timer running Case 1: - should not see Event 19 In transition from Active to Open Confirm, the connection must have a TCP connection completed. Case 1 does not have this occurring, so the transition must be avoided. Case 2: - should see Event 19 - Open, Keepalive should be sent. Previous text: (Action H from FSM document) If an Open is received with the BGP Delay Open timer running, [Event 19], the local system: - clears the connect retry timer [cleared to zero), - completes the BGP initialization, - stop and clears the BGP Open Delay timer, - Sends an Open Message, - sets the hold timer to a large value (4 minutes), and - changes its state to an Open Confirm. New text: [a New Action - N-2 : N + BGP keepalive sent] If an Open is received with the BGP Delay Open Timer running [Event 19], the local system: - **clears the connect retry timer [cleared to zero], - completes the BGP initialization, - stops and clears the BGP Open Delay timer, - **Sends an Open message, - Sends a Keepalive message, - If hold timer value is non-zero, - **sets keepalive timer - hold timer reset to negotiated value else if hold timer is zero, - **resets the keepalive timer, and - **resets the hold timer. ** and changes state to OpenConfirm - If the value of the autonomous system field is the same as the local Autonomous system number, set the connection status to a internal connection; otherwise it is "external". Please confirm that this change is OK for this comment. Sue -----Original Message----- From: Tom Petch [mailto:nwnetworks@dial.pipex.com] Sent: Thursday, November 28, 2002 9:54 AM To: idr Subject: bgp18 WG Last Call fsm missing keepalive When the OpenConfirm state is entered from OpenSent with the receipt of a valid open [Event 18], then a KeepAlive message is sent and the timer is started. When the OpenConfirm state is entered from Active or Connect on receipt of a valid open [Event 19], no message is sent, no timer is started. I believe this inconsistency is an error and should be corrected by adding these two actions in those two places. Tom Petch nwnetworks@dial.pipex.com Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id NAA22364 for <idr-archive@nic.merit.edu>; Wed, 4 Dec 2002 13:32:09 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 1D37A912BD; Wed, 4 Dec 2002 13:30:46 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id DEE13912BE; Wed, 4 Dec 2002 13:30:45 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 65A2F912BD for <idr@trapdoor.merit.edu>; Wed, 4 Dec 2002 13:30:44 -0500 (EST) Received: by segue.merit.edu (Postfix) id 3F8A35E0DD; Wed, 4 Dec 2002 13:30:42 -0500 (EST) Delivered-To: idr@merit.edu Received: from nemesis.systems.pipex.net (nemesis.systems.pipex.net [62.241.160.8]) by segue.merit.edu (Postfix) with ESMTP id 90F115DE20 for <idr@merit.edu>; Wed, 4 Dec 2002 13:30:41 -0500 (EST) Received: from tom3 (usercb13.uk.uudial.com [62.188.150.180]) by nemesis.systems.pipex.net (Postfix) with SMTP id 0125116007F9C; Wed, 4 Dec 2002 18:30:36 +0000 (GMT) Message-ID: <010601c29bc3$382a00c0$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: "Susan Hares" <shares@nexthop.com>, "idr" <idr@merit.edu> Cc: "Jeffrey Haas" <jhaas@nexthop.com> Subject: Re: bgp18 WG Last Call fsm missing keepalive Date: Wed, 4 Dec 2002 18:29:49 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Still not quite. This can occur in the Connect state as well as Active, when a TCP connection completes with delay open flag set, my putative OpenDelay state (sigh, point taken). And we should still go to OpenConfirm state at the bottom of the action list as is currently part of list H. Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Susan Hares <shares@nexthop.com> To: Tom Petch <nwnetworks@dial.pipex.com>; idr <idr@merit.edu> Cc: Jeffrey Haas <jhaas@nexthop.com> Date: 04 December 2002 18:07 Subject: RE: bgp18 WG Last Call fsm missing keepalive Tom: Just to clarify this comment: Event 19 = valid open with delay timer running **Connect or Active = 1) awaiting TCP connection, or 2) TCP connection completed and awaiting the TCP connection with delay timer running Case 1: - should not see Event 19 In transition from Active to Open Confirm, the connection must have a TCP connection completed. Case 1 does not have this occurring, so the transition must be avoided. Case 2: - should see Event 19 - Open, Keepalive should be sent. Previous text: (Action H from FSM document) If an Open is received with the BGP Delay Open timer running, [Event 19], the local system: - clears the connect retry timer [cleared to zero), - completes the BGP initialization, - stop and clears the BGP Open Delay timer, - Sends an Open Message, - sets the hold timer to a large value (4 minutes), and - changes its state to an Open Confirm. New text: [a New Action - N-2 : N + BGP keepalive sent] If an Open is received with the BGP Delay Open Timer running [Event 19], the local system: - **clears the connect retry timer [cleared to zero], - completes the BGP initialization, - stops and clears the BGP Open Delay timer, - **Sends an Open message, - Sends a Keepalive message, - If hold timer value is non-zero, - **sets keepalive timer - hold timer reset to negotiated value else if hold timer is zero, - **resets the keepalive timer, and - **resets the hold timer. ** and changes state to OpenConfirm - If the value of the autonomous system field is the same as the local Autonomous system number, set the connection status to a internal connection; otherwise it is "external". Please confirm that this change is OK for this comment. Sue -----Original Message----- From: Tom Petch [mailto:nwnetworks@dial.pipex.com] Sent: Thursday, November 28, 2002 9:54 AM To: idr Subject: bgp18 WG Last Call fsm missing keepalive When the OpenConfirm state is entered from OpenSent with the receipt of a valid open [Event 18], then a KeepAlive message is sent and the timer is started. When the OpenConfirm state is entered from Active or Connect on receipt of a valid open [Event 19], no message is sent, no timer is started. I believe this inconsistency is an error and should be corrected by adding these two actions in those two places. Tom Petch nwnetworks@dial.pipex.com Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id NAA21309 for <idr-archive@nic.merit.edu>; Wed, 4 Dec 2002 13:07:56 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 32064912BC; Wed, 4 Dec 2002 13:07:28 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id F3B54912BD; Wed, 4 Dec 2002 13:07:27 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 8E772912BC for <idr@trapdoor.merit.edu>; Wed, 4 Dec 2002 13:07:26 -0500 (EST) Received: by segue.merit.edu (Postfix) id 74C365E0DD; Wed, 4 Dec 2002 13:07:26 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id CA1525E0DA for <idr@merit.edu>; Wed, 4 Dec 2002 13:07:25 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB4I7OR30577 for idr@merit.edu; Wed, 4 Dec 2002 13:07:24 -0500 (EST) (envelope-from shares@nexthop.com) Received: from mail.corp.nexthop.com ([64.211.218.233]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB4I7JC30564 for <idr@merit.edu>; Wed, 4 Dec 2002 13:07:19 -0500 (EST) (envelope-from shares@nexthop.com) content-class: urn:content-classes:message Subject: RE: bgp18 WG Last Call fsm missing keepalive Date: Wed, 4 Dec 2002 13:07:19 -0500 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Message-ID: <BE36D687C8CBFA4AB76F5CD3E3C0FB4EA6AAF8@aa-exchange1.corp.nexthop.com> X-MS-Has-Attach: X-MIMEOLE: Produced By Microsoft Exchange V6.0.6249.0 X-MS-TNEF-Correlator: Thread-Topic: bgp18 WG Last Call fsm missing keepalive Thread-Index: AcKW7j4aVoADIlDnRJmySTnzE0qoPgEzTRsw From: "Susan Hares" <shares@nexthop.com> To: "Tom Petch" <nwnetworks@dial.pipex.com>, "idr" <idr@merit.edu> Cc: "Jeffrey Haas" <jhaas@nexthop.com> X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by nic.merit.edu id NAA21309 Tom: Just to clarify this comment: Event 19 = valid open with delay timer running Active = 1) awaiting TCP connection, or 2) TCP connection completed and awaiting the TCP connection with delay timer running Case 1: - should not see Event 19 In transition from Active to Open Confirm, the connection must have a TCP connection completed. Case 1 does not have this occurring, so the transition must be avoided. Case 2: - should see Event 19 - Open, Keepalive should be sent. Previous text: (Action H from FSM document) If an Open is received with the BGP Delay Open timer running, [Event 19], the local system: - clears the connect retry timer [cleared to zero), - completes the BGP initialization, - stop and clears the BGP Open Delay timer, - Sends an Open Message, - sets the hold timer to a large value (4 minutes), and - changes its state to an Open Confirm. New text: [a New Action - N-2 : N + BGP keepalive sent] If an Open is received with the BGP Delay Open Timer running [Event 19], the local system: - clear the connect retry timer [cleared to zero], - completes the BGP initialization, - stops and clears the BGP Open Delay timer, - Send an Open message, - Sends a Keepalive message, - If hold timer value is non-zero, - set keepalive timer - hold timer reset to negotiated value else if hold timer is zero, - reset the keepalive timer, and - reset the hold timer. - If the value of the autonomous system field is the same as the local Autonomous system number, set the connection status to a internal connection; otherwise it is "external". Please confirm that this change is OK for this comment. Sue -----Original Message----- From: Tom Petch [mailto:nwnetworks@dial.pipex.com] Sent: Thursday, November 28, 2002 9:54 AM To: idr Subject: bgp18 WG Last Call fsm missing keepalive When the OpenConfirm state is entered from OpenSent with the receipt of a valid open [Event 18], then a KeepAlive message is sent and the timer is started. When the OpenConfirm state is entered from Active or Connect on receipt of a valid open [Event 19], no message is sent, no timer is started. I believe this inconsistency is an error and should be corrected by adding these two actions in those two places. Tom Petch nwnetworks@dial.pipex.com Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id MAA19627 for <idr-archive@nic.merit.edu>; Wed, 4 Dec 2002 12:25:38 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 5D1B2912BB; Wed, 4 Dec 2002 12:25:19 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 2CA7F912BC; Wed, 4 Dec 2002 12:25:19 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id D26E1912BB for <idr@trapdoor.merit.edu>; Wed, 4 Dec 2002 12:25:16 -0500 (EST) Received: by segue.merit.edu (Postfix) id C19D65E0D3; Wed, 4 Dec 2002 12:25:16 -0500 (EST) Delivered-To: idr@merit.edu Received: from merlot.juniper.net (natint.juniper.net [207.17.136.129]) by segue.merit.edu (Postfix) with ESMTP id 383745DDD5 for <idr@merit.edu>; Wed, 4 Dec 2002 12:25:16 -0500 (EST) Received: from juniper.net (garnet.juniper.net [172.17.28.17]) by merlot.juniper.net (8.11.3/8.11.3) with ESMTP id gB4HPDS55420; Wed, 4 Dec 2002 09:25:13 -0800 (PST) (envelope-from yakov@juniper.net) Message-Id: <200212041725.gB4HPDS55420@merlot.juniper.net> To: "Tom Petch" <nwnetworks@dial.pipex.com> Cc: "idr" <idr@merit.edu> Subject: Re: bgp18 WG Last Call fsm incorrect next state In-Reply-To: Your message of "Thu, 28 Nov 2002 14:52:59 GMT." <002501c296ee$1b8e7920$0301a8c0@tom3> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <78814.1039022713.1@juniper.net> Date: Wed, 04 Dec 2002 09:25:13 -0800 From: Yakov Rekhter <yakov@juniper.net> Sender: owner-idr@merit.edu Precedence: bulk Tom, > I believe that there is an incorrect next state when the delay open > timer expires [event 12] in the Active state. The next state should > be OpenSent and not OpenConfirm. you are correct. Will be fixed in the next revision. Yakov. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id IAA21590 for <idr-archive@nic.merit.edu>; Tue, 3 Dec 2002 08:04:23 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 322DC91297; Tue, 3 Dec 2002 08:03:53 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id B0F34912A0; Tue, 3 Dec 2002 08:03:40 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 6AE8B91297 for <idr@trapdoor.merit.edu>; Tue, 3 Dec 2002 08:02:59 -0500 (EST) Received: by segue.merit.edu (Postfix) id 45FEC5E027; Tue, 3 Dec 2002 08:02:59 -0500 (EST) Delivered-To: idr@merit.edu Received: from ietf.org (odin.ietf.org [132.151.1.176]) by segue.merit.edu (Postfix) with ESMTP id 587F05DF39 for <idr@merit.edu>; Tue, 3 Dec 2002 08:02:58 -0500 (EST) Received: from CNRI.Reston.VA.US (localhost [127.0.0.1]) by ietf.org (8.9.1a/8.9.1a) with ESMTP id IAA17205; Tue, 3 Dec 2002 08:00:07 -0500 (EST) Message-Id: <200212031300.IAA17205@ietf.org> Mime-Version: 1.0 Content-Type: Multipart/Mixed; Boundary="NextPart" To: IETF-Announce: ; Cc: idr@merit.edu From: Internet-Drafts@ietf.org Reply-To: Internet-Drafts@ietf.org Subject: I-D ACTION:draft-ietf-idr-route-filter-07.txt Date: Tue, 03 Dec 2002 08:00:07 -0500 Sender: owner-idr@merit.edu Precedence: bulk --NextPart A New Internet-Draft is available from the on-line Internet-Drafts directories. This draft is a work item of the Inter-Domain Routing Working Group of the IETF. Title : Cooperative Route Filtering Capability for BGP-4 Author(s) : E. Chen, Y. Rekhter Filename : draft-ietf-idr-route-filter-07.txt Pages : 11 Date : 2002-12-2 This document defines a BGP-based mechanism that allows a BGP speaker to send to its BGP peer a set of route filters that the peer would use to constrain/filter its outbound routing updates to the speaker. A URL for this Internet-Draft is: http://www.ietf.org/internet-drafts/draft-ietf-idr-route-filter-07.txt To remove yourself from the IETF Announcement list, send a message to ietf-announce-request with the word unsubscribe in the body of the message. Internet-Drafts are also available by anonymous FTP. Login with the username "anonymous" and a password of your e-mail address. After logging in, type "cd internet-drafts" and then "get draft-ietf-idr-route-filter-07.txt". A list of Internet-Drafts directories can be found in http://www.ietf.org/shadow.html or ftp://ftp.ietf.org/ietf/1shadow-sites.txt Internet-Drafts can also be obtained by e-mail. Send a message to: mailserv@ietf.org. In the body type: "FILE /internet-drafts/draft-ietf-idr-route-filter-07.txt". NOTE: The mail server at ietf.org can return the document in MIME-encoded form by using the "mpack" utility. To use this feature, insert the command "ENCODING mime" before the "FILE" command. To decode the response(s), you will need "munpack" or a MIME-compliant mail reader. Different MIME-compliant mail readers exhibit different behavior, especially when dealing with "multipart" MIME messages (i.e. documents which have been split up into multiple messages), so check your local documentation on how to manipulate these messages. Below is the data which will enable a MIME compliant mail reader implementation to automatically retrieve the ASCII version of the Internet-Draft. --NextPart Content-Type: Multipart/Alternative; Boundary="OtherAccess" --OtherAccess Content-Type: Message/External-body; access-type="mail-server"; server="mailserv@ietf.org" Content-Type: text/plain Content-ID: <2002-12-2140100.I-D@ietf.org> ENCODING mime FILE /internet-drafts/draft-ietf-idr-route-filter-07.txt --OtherAccess Content-Type: Message/External-body; name="draft-ietf-idr-route-filter-07.txt"; site="ftp.ietf.org"; access-type="anon-ftp"; directory="internet-drafts" Content-Type: text/plain Content-ID: <2002-12-2140100.I-D@ietf.org> --OtherAccess-- --NextPart-- Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id IAA21388 for <idr-archive@nic.merit.edu>; Tue, 3 Dec 2002 08:00:30 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 1E4C491294; Tue, 3 Dec 2002 08:00:00 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id D3B3F91295; Tue, 3 Dec 2002 07:59:59 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 8F4C291294 for <idr@trapdoor.merit.edu>; Tue, 3 Dec 2002 07:59:58 -0500 (EST) Received: by segue.merit.edu (Postfix) id 6D05B5E0B3; Tue, 3 Dec 2002 07:59:58 -0500 (EST) Delivered-To: idr@merit.edu Received: from workhorse.fictitious.org (workhorse.fictitious.org [209.150.1.230]) by segue.merit.edu (Postfix) with ESMTP id B03975E008 for <idr@merit.edu>; Tue, 3 Dec 2002 07:59:57 -0500 (EST) Received: from workhorse.fictitious.org (localhost.fictitious.org [127.0.0.1]) by workhorse.fictitious.org (8.9.3/8.9.3) with ESMTP id HAA01283; Tue, 3 Dec 2002 07:59:20 -0500 (EST) (envelope-from curtis@workhorse.fictitious.org) Message-Id: <200212031259.HAA01283@workhorse.fictitious.org> To: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com> Cc: Mathew Richardson <mrr@nexthop.com>, idr@merit.edu Reply-To: curtis@fictitious.org Subject: Re: proxy: more comments on draft -18 In-reply-to: Your message of "Mon, 02 Dec 2002 23:23:07 +0530." <55E277B99171E041ABF5F4B1C6DDCA06DD2974@haritha.hclt.com> Date: Tue, 03 Dec 2002 07:59:20 -0500 From: Curtis Villamizar <curtis@fictitious.org> Sender: owner-idr@merit.edu Precedence: bulk In message <55E277B99171E041ABF5F4B1C6DDCA06DD2974@haritha.hclt.com>, "Sivanand a Ramnath - CTD, Chennai." writes: > Hello, > > That does not quite solve the problem. Curtis wrote a detailed mail on > the issue in which he explained a possibility for formation of loops if this > were done. > > Sivanand > (siva@ctd.hcltech.com) > > > > I've always thought (perhaps erroneously) that the reason support for > > removing MEDs was included in the spec was to allow an AS to choose to > > ignore MEDs. If it is desirable to allow an AS to choose to > > ignore MEDs, > > and this was the motivation for supporting the removal of MEDs, then I > > propose: > > > > If a MULTI_EXIT_DISC attribute will be removed before re- > > advertising a route into IBGP, then the attribute SHALL be > > considered missing when performing comparisons based on the > > MULTI_EXIT_DISC attribute of a route. > > > > mrr The text above would be sufficient to eliminate the route loops but the action recommended by it would be more than is needed to eliminate route loops and does not match current practice (afaik). It is safe to use the MED in comparisons of EBGP learned routes, pick one, then strip MED, then compare the best EBGP learned routes with the IBGP learned routes, then advertise. The text above oversimplifies yielding two (not quite correct) steps: strip MED, then compare. Curtis Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id WAA21370 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 22:49:06 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 4FFB591284; Mon, 2 Dec 2002 22:48:44 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 1B87191285; Mon, 2 Dec 2002 22:48:44 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id D137491284 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 22:48:42 -0500 (EST) Received: by segue.merit.edu (Postfix) id B49C95DF81; Mon, 2 Dec 2002 22:48:42 -0500 (EST) Delivered-To: idr@merit.edu Received: from ganesh.ctd.hctech.com (unknown [202.54.64.2]) by segue.merit.edu (Postfix) with ESMTP id 3F81B5DFDB for <idr@merit.edu>; Mon, 2 Dec 2002 22:48:41 -0500 (EST) Received: by GANESH with Internet Mail Service (5.5.2653.19) id <YD7Z6TPM>; Mon, 2 Dec 2002 23:29:12 +0530 Message-ID: <55E277B99171E041ABF5F4B1C6DDCA06DD2974@haritha.hclt.com> From: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com> To: Mathew Richardson <mrr@nexthop.com>, idr@merit.edu Subject: RE: proxy: more comments on draft -18 Date: Mon, 2 Dec 2002 23:23:07 +0530 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-idr@merit.edu Precedence: bulk Hello, That does not quite solve the problem. Curtis wrote a detailed mail on the issue in which he explained a possibility for formation of loops if this were done. Sivanand (siva@ctd.hcltech.com) > > I've always thought (perhaps erroneously) that the reason support for > removing MEDs was included in the spec was to allow an AS to choose to > ignore MEDs. If it is desirable to allow an AS to choose to > ignore MEDs, > and this was the motivation for supporting the removal of MEDs, then I > propose: > > If a MULTI_EXIT_DISC attribute will be removed before re- > advertising a route into IBGP, then the attribute SHALL be > considered missing when performing comparisons based on the > MULTI_EXIT_DISC attribute of a route. > > mrr > Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id TAA10120 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 19:24:24 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id DA1A491283; Mon, 2 Dec 2002 19:23:07 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id A762691282; Mon, 2 Dec 2002 19:23:07 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 7EBAD91281 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 19:23:04 -0500 (EST) Received: by segue.merit.edu (Postfix) id 53D7D5E016; Mon, 2 Dec 2002 19:23:04 -0500 (EST) Delivered-To: idr@merit.edu Received: from nomad.tcb.net (vdsl-151-118-3-177.dnvr.uswest.net [151.118.3.177]) by segue.merit.edu (Postfix) with ESMTP id 138105DFFD for <idr@merit.edu>; Mon, 2 Dec 2002 19:23:04 -0500 (EST) Received: by nomad.tcb.net (Postfix, from userid 500) id D0B5755F67; Mon, 2 Dec 2002 17:23:23 -0700 (MST) Received: from nomad.tcb.net (localhost [127.0.0.1]) by nomad.tcb.net (Postfix) with ESMTP id CDD5F3E83 for <idr@merit.edu>; Mon, 2 Dec 2002 17:23:23 -0700 (MST) X-Mailer: exmh version 2.5 07/13/2001 with nmh-1.0.4 To: idr@merit.edu From: Danny McPherson <danny@tcb.net> Reply-To: danny@tcb.net Subject: RFC1773bis - Experience with BGP Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 02 Dec 2002 17:23:18 -0700 Message-Id: <20021203002323.D0B5755F67@nomad.tcb.net> Sender: owner-idr@merit.edu Precedence: bulk Folks, As required by RFC 1264 as part of the BGP-4 standardization process RFC 1773 (Experience with BGP) needs to be updated, and I've volunteered to do the update. Namely, this update needs to focus on [from RFC 1264]: Revisions to the Protocol and Usage documents showing changes and clarifications made based on experience gained in the time between when the protocol was made a Draft Standard and it being submitted for Standard. The changes should be to clarify the protocol or provide guidance in its implementation. With that said, (and 4 other items from 1264 being required as well), I'd like to solicit feedback from the WG on what items they'd like to see in the update, in addition to those items discussed in Appendix A of idr-bgp4-18.txt. Thanks in advance! -danny Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id SAA07470 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 18:34:03 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 083AD91238; Mon, 2 Dec 2002 18:33:36 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 8360D9127C; Mon, 2 Dec 2002 18:32:10 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 330D591238 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 18:32:08 -0500 (EST) Received: by segue.merit.edu (Postfix) id 067765DEA4; Mon, 2 Dec 2002 18:32:08 -0500 (EST) Delivered-To: idr@merit.edu Received: from demiurge.exodus.net (demiurge.exodus.net [216.32.171.82]) by segue.merit.edu (Postfix) with ESMTP id 6B91F5DDC5 for <idr@merit.edu>; Mon, 2 Dec 2002 18:32:07 -0500 (EST) Received: (from andrewl@localhost) by demiurge.exodus.net (8.9.3+Sun/8.9.3) id PAA21422; Mon, 2 Dec 2002 15:28:45 -0800 (PST) Date: Mon, 2 Dec 2002 15:28:45 -0800 From: andrewl@xix-w.bengi.exodus.net To: Yakov Rekhter <yakov@juniper.net> Cc: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com>, Alex Zinin <zinin@psg.com>, idr@merit.edu Subject: Re: proxy: more comments on draft -18 Message-ID: <20021202152845.A18698@demiurge.exodus.net> References: <55E277B99171E041ABF5F4B1C6DDCA06DB41CC@haritha.hclt.com> <200212021454.gB2EsYS39593@merlot.juniper.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <200212021454.gB2EsYS39593@merlot.juniper.net>; from yakov@juniper.net on Mon, Dec 02, 2002 at 06:54:34AM -0800 Sender: owner-idr@merit.edu Precedence: bulk > > > here is my attempt to clarify this: > > > > > > If a MULTI_EXIT_DISC attribute is removed before re-advertising a > > > route into IBGP, then (prior to the removal) the MULTI_EXIT_DISC > > > attribute may only be considered in the comparison of EBGP learned > > > routes; the attribute is then removed, and then the remaining EBGP > > > learned routes may be compared to the remaining IBGP learned routes, > > > without considering the MULTI_EXIT_DISC attribute for those EBGP > > > learned routes whose MULTI_EXIT_DISC attribute will be removed > > > before advertising these routes to IBGP. > > > > > > Any further suggestions on how to improve this would be appreciated. > > > > How about this: > > > > % If a MULTI_EXIT_DISC attribute is removed before re-advertising a > > % route into IBGP, then comparison based on the MULT_EXIT_DISC > > % attribute may (MUST?) be performed only among the EBGP learned routes. > > % This comparison MUST be performed before the removal of the > > % MULTI_EXIT_DISC attribute. The MULT_EXIT_DISC attribute must then > > % be removed from those EBGP routes where such removal is required and > > % which are still eligible. This is followed by comparison with IBGP > > % learned routes. > > > > I think this reflects our objectives, which is: > > > > a) If MED is to be removed, compare EBGP routes based on the MED > > > > b) Then remove the MED > > > > c) Then do comparison with IBGP routes > > If nobody would propose anything better, then I'll use your text in > the next revision. How about: If a router is configured to remove a MUTLI_EXIT_DISC attribute from a route learned from EBGP, before readvertising it into IBGP the router MUST compare the route with other EBGP-learned routes before removing the MULTI_EXIT_DISC. Once this comparison is complete, the MED may be removed, and any remaining routes can be compared with IBGP routes to determine the best route. Andrew Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id OAA28637 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 14:49:03 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 2B2E391275; Mon, 2 Dec 2002 14:47:48 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id F300191276; Mon, 2 Dec 2002 14:47:47 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 6DA7A91275 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 14:47:23 -0500 (EST) Received: by segue.merit.edu (Postfix) id 3C80C5DE4B; Mon, 2 Dec 2002 14:47:23 -0500 (EST) Delivered-To: idr@merit.edu Received: from merlot.juniper.net (natint.juniper.net [207.17.136.129]) by segue.merit.edu (Postfix) with ESMTP id 649625E021 for <idr@merit.edu>; Mon, 2 Dec 2002 14:47:18 -0500 (EST) Received: from juniper.net (garnet.juniper.net [172.17.28.17]) by merlot.juniper.net (8.11.3/8.11.3) with ESMTP id gB2JlIS72581 for <idr@merit.edu>; Mon, 2 Dec 2002 11:47:18 -0800 (PST) (envelope-from yakov@juniper.net) Message-Id: <200212021947.gB2JlIS72581@merlot.juniper.net> To: idr@merit.edu Subject: Close of Last Call MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <40107.1038858437.1@juniper.net> Date: Mon, 02 Dec 2002 11:47:17 -0800 From: Yakov Rekhter <yakov@juniper.net> Sender: owner-idr@merit.edu Precedence: bulk Folks, The last call on BGP Base spec: draft-ietf-idr-bgp4-18.txt has ended. The editors will reply to all the comments received so far and will re-issue a revised draft that reflects the comments. Yakov. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id OAA28410 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 14:44:40 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 1AB9591270; Mon, 2 Dec 2002 14:44:21 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id D66D591273; Mon, 2 Dec 2002 14:44:20 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id BE3A991270 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 14:44:19 -0500 (EST) Received: by segue.merit.edu (Postfix) id A58E95E021; Mon, 2 Dec 2002 14:44:19 -0500 (EST) Delivered-To: idr@merit.edu Received: from funnel.cisco.com (funnel.cisco.com [161.44.168.79]) by segue.merit.edu (Postfix) with ESMTP id 1111A5DE4B for <idr@merit.edu>; Mon, 2 Dec 2002 14:44:19 -0500 (EST) Received: from bifocal.cisco.com (bifocal.cisco.com [161.44.172.150]) by funnel.cisco.com (8.8.5-Cisco.1/8.6.5) with ESMTP id OAA06992; Mon, 2 Dec 2002 14:44:17 -0500 (EST) Received: from localhost (swallow@localhost) by bifocal.cisco.com (8.8.8-Cisco List Logging/CISCO.WS.1.2) with ESMTP id OAA19839; Mon, 2 Dec 2002 14:44:17 -0500 (EST) Message-Id: <200212021944.OAA19839@bifocal.cisco.com> X-Authentication-Warning: bifocal.cisco.com: swallow owned process doing -bs To: idr@merit.edu, mpls@uu.net Subject: Close of Last Call Date: Mon, 02 Dec 2002 14:44:17 -0500 From: George Swallow <swallow@cisco.com> Sender: owner-idr@merit.edu Precedence: bulk The last call on "Graceful Restart Mechanisms for BGP with MPLS" draft-ietf-mpls-bgp-mpls-restart-02.txt has ended. ...George ====================================================================== George Swallow Cisco Systems (978) 497-8143 250 Apollo Drive Chelmsford, Ma 01824 Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id MAA23520 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 12:42:07 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id B7BCE91262; Mon, 2 Dec 2002 12:41:48 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 7D45F91263; Mon, 2 Dec 2002 12:41:48 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 3459391262 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 12:41:47 -0500 (EST) Received: by segue.merit.edu (Postfix) id 187705DEB2; Mon, 2 Dec 2002 12:41:47 -0500 (EST) Delivered-To: idr@merit.edu Received: from merlot.juniper.net (natint.juniper.net [207.17.136.129]) by segue.merit.edu (Postfix) with ESMTP id 7EFCE5DE89 for <idr@merit.edu>; Mon, 2 Dec 2002 12:41:46 -0500 (EST) Received: from juniper.net (garnet.juniper.net [172.17.28.17]) by merlot.juniper.net (8.11.3/8.11.3) with ESMTP id gB2HfkS59895 for <idr@merit.edu>; Mon, 2 Dec 2002 09:41:46 -0800 (PST) (envelope-from yakov@juniper.net) Message-Id: <200212021741.gB2HfkS59895@merlot.juniper.net> To: idr@merit.edu Subject: slides MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <3016.1038850906.1@juniper.net> Date: Mon, 02 Dec 2002 09:41:46 -0800 From: Yakov Rekhter <yakov@juniper.net> Sender: owner-idr@merit.edu Precedence: bulk Folks, Those of you who presented at the last IDR meeting, please send your slides to minutes@ietf.org Yakov. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id MAA22747 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 12:24:15 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id A617991261; Mon, 2 Dec 2002 12:23:52 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 6781491262; Mon, 2 Dec 2002 12:23:52 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 2EDE091261 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 12:23:51 -0500 (EST) Received: by segue.merit.edu (Postfix) id 1251E5DE89; Mon, 2 Dec 2002 12:23:51 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 77BF05DDE2 for <idr@merit.edu>; Mon, 2 Dec 2002 12:23:50 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB2HNnb77605 for idr@merit.edu; Mon, 2 Dec 2002 12:23:49 -0500 (EST) (envelope-from mrr@mrichardson.nexthop.com) Received: from paradigm.nexthop.com (mrichardson.nexthop.com [64.211.218.45]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB2HNkC77598 for <idr@merit.edu>; Mon, 2 Dec 2002 12:23:46 -0500 (EST) (envelope-from mrr@mrichardson.nexthop.com) Received: (from mrr@localhost) by paradigm.nexthop.com (8.11.6/8.11.6) id gB2HNkw18921 for idr@merit.edu; Mon, 2 Dec 2002 12:23:46 -0500 (EST) Date: Mon, 2 Dec 2002 12:23:46 -0500 From: Mathew Richardson <mrr@nexthop.com> To: idr@merit.edu Subject: Re: proxy: more comments on draft -18 Message-ID: <20021202172346.GB17400@nexthop.com> References: <55E277B99171E041ABF5F4B1C6DDCA06DB41CC@haritha.hclt.com> <200212021454.gB2EsYS39593@merlot.juniper.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <200212021454.gB2EsYS39593@merlot.juniper.net> User-Agent: Mutt/1.4i X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk > Yakov Rekhter <yakov@juniper.net> [Mon, Dec 02, 2002 at 06:54:34AM -0800]: >Siva, <snip> >> > here is my attempt to clarify this: >> > >> > If a MULTI_EXIT_DISC attribute is removed before re-advertising a >> > route into IBGP, then (prior to the removal) the MULTI_EXIT_DISC >> > attribute may only be considered in the comparison of EBGP learned >> > routes; the attribute is then removed, and then the remaining EBGP >> > learned routes may be compared to the remaining IBGP learned routes, >> > without considering the MULTI_EXIT_DISC attribute for those EBGP >> > learned routes whose MULTI_EXIT_DISC attribute will be removed >> > before advertising these routes to IBGP. >> > >> > Any further suggestions on how to improve this would be appreciated. >> >> How about this: >> >> % If a MULTI_EXIT_DISC attribute is removed before re-advertising a >> % route into IBGP, then comparison based on the MULT_EXIT_DISC >> % attribute may (MUST?) be performed only among the EBGP learned routes. >> % This comparison MUST be performed before the removal of the >> % MULTI_EXIT_DISC attribute. The MULT_EXIT_DISC attribute must then >> % be removed from those EBGP routes where such removal is required and >> % which are still eligible. This is followed by comparison with IBGP >> % learned routes. >> >> I think this reflects our objectives, which is: >> >> a) If MED is to be removed, compare EBGP routes based on the MED >> >> b) Then remove the MED >> >> c) Then do comparison with IBGP routes > >If nobody would propose anything better, then I'll use your text in >the next revision. <snip> I've always thought (perhaps erroneously) that the reason support for removing MEDs was included in the spec was to allow an AS to choose to ignore MEDs. If it is desirable to allow an AS to choose to ignore MEDs, and this was the motivation for supporting the removal of MEDs, then I propose: If a MULTI_EXIT_DISC attribute will be removed before re- advertising a route into IBGP, then the attribute SHALL be considered missing when performing comparisons based on the MULTI_EXIT_DISC attribute of a route. mrr Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id LAA20621 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 11:36:12 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 6973C9125F; Mon, 2 Dec 2002 11:33:34 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 79B6891261; Mon, 2 Dec 2002 11:33:31 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id C6BA791260 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 11:33:27 -0500 (EST) Received: by segue.merit.edu (Postfix) id 3B4855DDDF; Mon, 2 Dec 2002 11:33:27 -0500 (EST) Delivered-To: idr@merit.edu Received: from smail2.alcatel.fr (gc-na5.alcatel.fr [64.208.49.5]) by segue.merit.edu (Postfix) with ESMTP id 92AF75DDD2 for <idr@merit.edu>; Mon, 2 Dec 2002 11:33:26 -0500 (EST) Received: from nmu.alcatel.fr (tcmh80.nmu.alcatel.fr [139.54.143.3]) by smail2.alcatel.fr (ALCANET/NETFR) with ESMTP id gB2GXPaX011915 for <idr@merit.edu>; Mon, 2 Dec 2002 17:33:25 +0100 Received: from nmu.alcatel.fr (houat [192.200.245.153]) by nmu.alcatel.fr (8.8.6 (PHNE_17135)/8.8.6) with ESMTP id RAA05425 for <idr@merit.edu>; Mon, 2 Dec 2002 17:33:06 +0100 (MET) Message-ID: <3DEB8B43.251AFD8B@nmu.alcatel.fr> Date: Mon, 02 Dec 2002 17:33:07 +0100 From: christophe preguica <Christophe.Preguica@alcatel.fr> X-Mailer: Mozilla 4.79 [en] (Windows NT 5.0; U) X-Accept-Language: en MIME-Version: 1.0 To: idr@merit.edu Subject: MD5 and MP-BGP Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Virus-Scanned: by amavisd-milter (http://amavis.org/) Sender: owner-idr@merit.edu Precedence: bulk Does anyone know if TCP MD5 signature option (RFC 2385) for MP-BGP is applicable for IPv6 ? Thanks. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id LAA20358 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 11:27:46 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id A7BEE9125D; Mon, 2 Dec 2002 11:27:20 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 6F55F9125E; Mon, 2 Dec 2002 11:27:20 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 634549125D for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 11:27:19 -0500 (EST) Received: by segue.merit.edu (Postfix) id 4EB5F5DF37; Mon, 2 Dec 2002 11:27:19 -0500 (EST) Delivered-To: idr@merit.edu Received: from sentry.gw.tislabs.com (sentry.gw.tislabs.com [192.94.214.100]) by segue.merit.edu (Postfix) with ESMTP id CAC185DDDF for <idr@merit.edu>; Mon, 2 Dec 2002 11:27:18 -0500 (EST) Received: by sentry.gw.tislabs.com; id LAA08767; Mon, 2 Dec 2002 11:27:32 -0500 (EST) Received: from raven.gw.tislabs.com(10.33.1.50) by sentry.gw.tislabs.com via smap (V5.5) id xma008754; Mon, 2 Dec 02 11:26:58 -0500 Received: (from sandy@localhost) by raven.gw.tislabs.com (8.11.6/8.11.6) id gB2GQhk21933; Mon, 2 Dec 2002 11:26:43 -0500 (EST) Date: Mon, 2 Dec 2002 11:26:43 -0500 (EST) Message-Id: <200212021626.gB2GQhk21933@raven.gw.tislabs.com> From: sandy@tislabs.com To: acb@barberair.com, idr@merit.edu, sandy@tislabs.com Subject: Re: BGP security in practice Sender: owner-idr@merit.edu Precedence: bulk >>In the judgement of those on this list, is the general ISP as clueless >>as this? Mind you, I'm not asking if the common ISP uses the MD5, just >>if they know that there's a security vulnerability if they don't. > > >They probably also know there is a gaping hole even if they do - re TCP RST >attacks etc. But the TCP MD5 protections were said to be a good idea expressly because they do protect against TCP RST. Do you have news to the contrary? --Sandy Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id LAA20155 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 11:23:07 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 8AFFB91230; Mon, 2 Dec 2002 11:22:45 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 4E9A59125D; Mon, 2 Dec 2002 11:22:45 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 8306B91230 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 11:22:41 -0500 (EST) Received: by segue.merit.edu (Postfix) id 7271E5DF37; Mon, 2 Dec 2002 11:22:41 -0500 (EST) Delivered-To: idr@merit.edu Received: from nemesis.systems.pipex.net (nemesis.systems.pipex.net [62.241.160.8]) by segue.merit.edu (Postfix) with ESMTP id 26CEB5DDDF for <idr@merit.edu>; Mon, 2 Dec 2002 11:22:41 -0500 (EST) Received: from tom3 (userbm24.uk.uudial.com [62.188.144.230]) by nemesis.systems.pipex.net (Postfix) with SMTP id 6BC9916007E5C; Mon, 2 Dec 2002 16:22:35 +0000 (GMT) Message-ID: <004b01c29a1f$033ba760$0301a8c0@tom3> Reply-To: "Tom Petch" <nwnetworks@dial.pipex.com> From: "Tom Petch" <nwnetworks@dial.pipex.com> To: "Jeffrey Haas" <jhaas@nexthop.com> Cc: "idr" <idr@merit.edu> Subject: Re: bgp18 WG Last Call fsm missing next state Date: Mon, 2 Dec 2002 16:06:30 -0000 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Sender: owner-idr@merit.edu Precedence: bulk Jeff I appreciate your comments, agreeing with many. A minor supplementary; Open Confirm, valid open [event 18], decide to drop this connection and so the FSM but what about the actions of incrementing the ConnectRetryCnt (and optional peer oscillation damping)? Are we doing that in the other FSM? on a per router id or per IP address pair independent of FSM? And a slightly bigger issue in the Active state (which is the passive state because we have entered it from Idle with Passive TCP flag set Events 4/5). In which case, TCP Indication valid [event 13] or invalid [event 14] are what I would expect with event 14 keeping us in Active state; and if event 13 also keeps us in Active state (while sending a TCP SYN-ACK) then TCP Connection Confirmed [event 16] will ensue while TCP Connection Acknowledged [event 15] would be an FSM error. Except that you suggest that event 14 in the Connect state takes us to Active state which seems fine but for the fact that we will have sent a TCP SYN (because that is what the Connect state is), this invalid Indication may be a coincidence, in which case our SYN is about to receive a SYN-ACK back ie event 15 is valid after all. And I am then in agreement with you over events 15/16 in the Active (ie passive) state (delay open flag keeps us in the Active state). Tom Petch nwnetworks@dial.pipex.com -----Original Message----- From: Jeffrey Haas <jhaas@nexthop.com> To: Tom Petch <nwnetworks@dial.pipex.com> Cc: idr <idr@merit.edu> Date: 30 November 2002 01:33 Subject: Re: bgp18 WG Last Call fsm missing next state >Tom, > >On Thu, Nov 28, 2002 at 02:53:56PM -0000, Tom Petch wrote: >> In seven places in the FSM, the actions listed for the occurrence of >> an event do not include any indication of what the next state should >> be, nor can I be sure I can guess what it should be. >> >> I believe the FSM should specify the new state in the cases listed >> below. >> >> Connect State: >> >> If the TCP connection succeeds [Event 15 or >> Event 16], the local system checks the "Delay Open >> Flag". If the delay Open flag is set, the local system: >> **enters what state > >Remains in the connect state awaiting either an open message or >for the opendelay timer to expire. Agreed > >> If the TCP connection receives an indication >> that is invalid or unconfigured. [Event 14]: >> **enters what state > >Enters the active state. >Connectretrytimer should be started. Agreed > >> Active State: >> >> A TCP connection succeeds [Event 15 or Event 16], the >> local system: process the TCP connection flags >> - If the BGP delay open flag is set: >> ** enters what state (I think this is an FSM error in TCP because it >> has not initiated a connection!) > >In the active state, we are awaiting incoming tcp connections to be >completed. Here we should: > >remain in the active state >wait for an open message or the opendelay timer to expire. > >> If the local system receives a valid TCP Indication >> [Event 13], the local system processes the TCP connection >> flags. >> ** enters what state > >I'm also confused by this event. Since we are not initiating a TCP >connection, but are awaiting one to complete, why would we get an >Indication here? FSM error? > >> If the local system receives a TCP indication >> that is invalid for this connection [Event 14]: >> ** enters what state > >Ditto. > >> >> If the local system receives a TCP connection >> failed [Event 17] (timeout or receives connection >> disconnect), the local system will: >> ** enters what state > >Ditto. > >> Open Confirm State: >> >> If the Open messages is valid [Event 18], the collision >> detect function is processed per section 6.8. If this >> connection is to be dropped due to call collision, the >> local system: >> ** enters what state > >This instance of the FSM should be discarded since it implies there is >another FSM instance running that will proceed to Established? > >> Tom Petch >> nwnetworks@dial.pipex.com > >-- >Jeff Haas >NextHop Technologies Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id LAA19945 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 11:18:20 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id B3F839125C; Mon, 2 Dec 2002 11:17:40 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 7D5D391258; Mon, 2 Dec 2002 11:17:40 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id BEFC79125C for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 11:17:38 -0500 (EST) Received: by segue.merit.edu (Postfix) id AB9565DF37; Mon, 2 Dec 2002 11:17:38 -0500 (EST) Delivered-To: idr@merit.edu Received: from merlot.juniper.net (natint.juniper.net [207.17.136.129]) by segue.merit.edu (Postfix) with ESMTP id 1B6355DDDF for <idr@merit.edu>; Mon, 2 Dec 2002 11:17:38 -0500 (EST) Received: from juniper.net (garnet.juniper.net [172.17.28.17]) by merlot.juniper.net (8.11.3/8.11.3) with ESMTP id gB2GHUS52750; Mon, 2 Dec 2002 08:17:30 -0800 (PST) (envelope-from yakov@juniper.net) Message-Id: <200212021617.gB2GHUS52750@merlot.juniper.net> To: "John G. Scudder" <jgs@cisco.com> Cc: idr@merit.edu Subject: Re: -18 last call comments In-Reply-To: Your message of "Mon, 02 Dec 2002 10:54:45 EST." <p05200f10ba112e474dfe@[192.168.42.10]> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <79651.1038845850.1@juniper.net> Date: Mon, 02 Dec 2002 08:17:30 -0800 From: Yakov Rekhter <yakov@juniper.net> Sender: owner-idr@merit.edu Precedence: bulk John, > At 10:25 AM -0500 12/2/02, Jeffrey Haas wrote: > >On Mon, Dec 02, 2002 at 01:03:50PM -0800, Manav wrote: > >> I understand that there can be situations when you receive your own AS > >> number in the UPDATE message and the best course of action is to simply > >> ignore it and proceed with the next set. > > > >In the case of receiving your own AS number, this is fairly typical. > > > >You may wish to search the archives for the thread on outbound > >as-loop detection. > [etc] > > I agree with Jeff's comments. > > I have one further comment just in case it's not perfectly obvious to > everyone, which is that "ignore the UPDATE" is not strictly the > action you take when receiving a looped update. Rather, you treat it > as an implicit withdraw, i.e. you process it as any other update but > treat the contained NLRI as unfeasible. > > I was going to write that this is sufficiently clear from the spec, > but I regret to say that it isn't. Here is the fourth paragraph of > section 9: > > The information carried by the AS_PATH attribute is checked for AS > loops. AS loop detection is done by scanning the full AS path (as > specified in the AS_PATH attribute), and checking that the autonomous > system number of the local system does not appear in the AS path. If > the autonomous system number appears in the AS path the route may be > stored in the Adj-RIB-In, but unless the router is configured to > accept routes with its own autonomous system in the AS path, the > route shall not be passed to the BGP Decision Process. Operations of > a router that is configured to accept routes with its own autonomous > system number in the AS path are outside the scope of this document. > > I don't think this is quite right -- the decision process needs to be > run if the looped routes had previously been advertised feasibly on > the same session. This could be fixed by hacking the quoted > paragraph, but it seems more straightforward to do it by removing the > quoted paragraph and making the fix in 9.1.2 Phase 2 instead. This > could be done by inserting the following between the third and fourth > paragraphs of 9.1.2 Phase 2: > > If the AS_PATH attribute of a BGP route contains an AS loop, the > BGP route should be excluded from the Phase 2 decision function. > AS loop detection is done by scanning the full AS path (as > specified in the AS_PATH attribute), and checking that the autonomous > system number of the local system does not appear in the AS path. > Operations of a router that is configured to accept routes with its > own autonomous system number in the AS path are outside the scope of > this document. > > Section 9.3, first bullet, also addresses this topic, but I don't > think it's sufficient. The change you suggested seems reasonable. Will be incorporated in the next revision. Yakov. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id LAA19463 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 11:07:09 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 22E4F9122F; Mon, 2 Dec 2002 11:06:36 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id DEA4D9125B; Mon, 2 Dec 2002 11:06:35 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id B21529122F for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 11:06:34 -0500 (EST) Received: by segue.merit.edu (Postfix) id A10205E087; Mon, 2 Dec 2002 11:06:34 -0500 (EST) Delivered-To: idr@merit.edu Received: from merlot.juniper.net (natint.juniper.net [207.17.136.129]) by segue.merit.edu (Postfix) with ESMTP id 127855E015 for <idr@merit.edu>; Mon, 2 Dec 2002 11:06:34 -0500 (EST) Received: from juniper.net (garnet.juniper.net [172.17.28.17]) by merlot.juniper.net (8.11.3/8.11.3) with ESMTP id gB2G6WS51976; Mon, 2 Dec 2002 08:06:32 -0800 (PST) (envelope-from yakov@juniper.net) Message-Id: <200212021606.gB2G6WS51976@merlot.juniper.net> To: Jeffrey Haas <jhaas@nexthop.com> Cc: idr@merit.edu Subject: Re: -18 last call comments In-Reply-To: Your message of "Mon, 02 Dec 2002 10:52:49 EST." <20021202105249.F14346@nexthop.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <76755.1038845192.1@juniper.net> Date: Mon, 02 Dec 2002 08:06:32 -0800 From: Yakov Rekhter <yakov@juniper.net> Sender: owner-idr@merit.edu Precedence: bulk Jeff, > On Sat, Nov 30, 2002 at 04:46:57PM -0800, Yakov Rekhter wrote: > > 1772 is in the non-normative section. I think we could keep it there, > > and keep the references as well. > > My concern is mostly that as we publish the BGP spec as an RFC > it'd be nice for it to refer to whatever replaces 1772. > > I don't know how this is handled in practice, but I don't think its > a big deal. I just thought it was worth bringing up. > > > > Capitalization of magic words such as MUST and SHOULD should be done. > > > > It should be done *as appropriate*. Therefore, if you think there > > are places where it is *appropriate* to do capitalization, please > > point them out. > > Will do when I can find a few cycles. > > > > Update message error subcode 7 is removed. Especially in -18, > > > it looks like an editing mistake based on where it would fall > > > in the editing.. > > > > From Appendix A: > > > > UPDATE Message Error subcode 7 (AS Routing Loop) has been deprecated. > > What I would like to see is something like this: > > 6 - Invalid ORIGIN Attribute > 7 - [Deprecated - See Appendix A] > 8 - Invalid NEXT_HOP Attribute > As it stands, 7 lies on a page boundary and looks like it got clipped > by the roff. Ok. I'll also add a similar text for Open Message Error subcode 5. Yakov. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id KAA19090 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 10:57:35 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id BB2EF91259; Mon, 2 Dec 2002 10:57:08 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 7CAFB9125A; Mon, 2 Dec 2002 10:57:08 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 76B7C91259 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 10:57:07 -0500 (EST) Received: by segue.merit.edu (Postfix) id 64CB25DDDF; Mon, 2 Dec 2002 10:57:07 -0500 (EST) Delivered-To: idr@merit.edu Received: from cisco.com (router.cisco.com [171.69.182.20]) by segue.merit.edu (Postfix) with ESMTP id AE51B5DDB8 for <idr@merit.edu>; Mon, 2 Dec 2002 10:57:06 -0500 (EST) Received: from [192.168.42.10] (dhcp-64-101-214-208.cisco.com [64.101.214.208]) by cisco.com (8.8.8/2.6/Cisco List Logging/8.8.8) with ESMTP id KAA19878 for <idr@merit.edu>; Mon, 2 Dec 2002 10:57:02 -0500 (EST) Mime-Version: 1.0 X-Sender: jgs@router Message-Id: <p05200f10ba112e474dfe@[192.168.42.10]> In-Reply-To: <20021202102540.A14346@nexthop.com> References: <00bf01c29a46$503abc50$8946d5a5@samsungy77z6fd> <20021202102540.A14346@nexthop.com> Date: Mon, 2 Dec 2002 10:54:45 -0500 To: idr@merit.edu From: "John G. Scudder" <jgs@cisco.com> Subject: Re: -18 last call comments Content-Type: text/plain; charset="us-ascii" ; format="flowed" Sender: owner-idr@merit.edu Precedence: bulk At 10:25 AM -0500 12/2/02, Jeffrey Haas wrote: >On Mon, Dec 02, 2002 at 01:03:50PM -0800, Manav wrote: >> I understand that there can be situations when you receive your own AS >> number in the UPDATE message and the best course of action is to simply >> ignore it and proceed with the next set. > >In the case of receiving your own AS number, this is fairly typical. > >You may wish to search the archives for the thread on outbound >as-loop detection. [etc] I agree with Jeff's comments. I have one further comment just in case it's not perfectly obvious to everyone, which is that "ignore the UPDATE" is not strictly the action you take when receiving a looped update. Rather, you treat it as an implicit withdraw, i.e. you process it as any other update but treat the contained NLRI as unfeasible. I was going to write that this is sufficiently clear from the spec, but I regret to say that it isn't. Here is the fourth paragraph of section 9: The information carried by the AS_PATH attribute is checked for AS loops. AS loop detection is done by scanning the full AS path (as specified in the AS_PATH attribute), and checking that the autonomous system number of the local system does not appear in the AS path. If the autonomous system number appears in the AS path the route may be stored in the Adj-RIB-In, but unless the router is configured to accept routes with its own autonomous system in the AS path, the route shall not be passed to the BGP Decision Process. Operations of a router that is configured to accept routes with its own autonomous system number in the AS path are outside the scope of this document. I don't think this is quite right -- the decision process needs to be run if the looped routes had previously been advertised feasibly on the same session. This could be fixed by hacking the quoted paragraph, but it seems more straightforward to do it by removing the quoted paragraph and making the fix in 9.1.2 Phase 2 instead. This could be done by inserting the following between the third and fourth paragraphs of 9.1.2 Phase 2: If the AS_PATH attribute of a BGP route contains an AS loop, the BGP route should be excluded from the Phase 2 decision function. AS loop detection is done by scanning the full AS path (as specified in the AS_PATH attribute), and checking that the autonomous system number of the local system does not appear in the AS path. Operations of a router that is configured to accept routes with its own autonomous system number in the AS path are outside the scope of this document. Section 9.3, first bullet, also addresses this topic, but I don't think it's sufficient. --John Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id KAA18803 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 10:53:17 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id DF9F291206; Mon, 2 Dec 2002 10:52:55 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id A96D291259; Mon, 2 Dec 2002 10:52:55 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 9F37391206 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 10:52:54 -0500 (EST) Received: by segue.merit.edu (Postfix) id 8262E5DDB8; Mon, 2 Dec 2002 10:52:54 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 1A3EC5E088 for <idr@merit.edu>; Mon, 2 Dec 2002 10:52:54 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB2Fqqt74572; Mon, 2 Dec 2002 10:52:52 -0500 (EST) (envelope-from jhaas@jhaas.nexthop.com) Received: from jhaas.nexthop.com (jhaas.nexthop.com [64.211.218.31]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB2FqnC74565; Mon, 2 Dec 2002 10:52:49 -0500 (EST) (envelope-from jhaas@jhaas.nexthop.com) Received: (from jhaas@localhost) by jhaas.nexthop.com (8.11.3nb1/8.11.3) id gB2Fqn314728; Mon, 2 Dec 2002 10:52:49 -0500 (EST) Date: Mon, 2 Dec 2002 10:52:49 -0500 From: Jeffrey Haas <jhaas@nexthop.com> To: Yakov Rekhter <yakov@juniper.net> Cc: idr@merit.edu Subject: Re: -18 last call comments Message-ID: <20021202105249.F14346@nexthop.com> References: <20021129200944.A10124@nexthop.com> <200212010046.gB10kvS75662@merlot.juniper.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <200212010046.gB10kvS75662@merlot.juniper.net>; from yakov@juniper.net on Sat, Nov 30, 2002 at 04:46:57PM -0800 X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Yakov, On Sat, Nov 30, 2002 at 04:46:57PM -0800, Yakov Rekhter wrote: > 1772 is in the non-normative section. I think we could keep it there, > and keep the references as well. My concern is mostly that as we publish the BGP spec as an RFC it'd be nice for it to refer to whatever replaces 1772. I don't know how this is handled in practice, but I don't think its a big deal. I just thought it was worth bringing up. > > Capitalization of magic words such as MUST and SHOULD should be done. > > It should be done *as appropriate*. Therefore, if you think there > are places where it is *appropriate* to do capitalization, please > point them out. Will do when I can find a few cycles. > > Update message error subcode 7 is removed. Especially in -18, > > it looks like an editing mistake based on where it would fall > > in the editing.. > > From Appendix A: > > UPDATE Message Error subcode 7 (AS Routing Loop) has been deprecated. What I would like to see is something like this: 6 - Invalid ORIGIN Attribute 7 - [Deprecated - See Appendix A] 8 - Invalid NEXT_HOP Attribute As it stands, 7 lies on a page boundary and looks like it got clipped by the roff. > > Reference to IDRP should be ISO10747 not IS10747. > > No, it should be IS10747. I'll take your word on this, however what I have on my shelf reads: ISO/IEC DIS 10747: 1992 Which probably means I have a last-call version of the draft standard. > Yakov. -- Jeff Haas NextHop Technologies Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id KAA17658 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 10:26:13 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 1169591257; Mon, 2 Dec 2002 10:25:51 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id D535691258; Mon, 2 Dec 2002 10:25:50 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id BB00591257 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 10:25:49 -0500 (EST) Received: by segue.merit.edu (Postfix) id A20325E019; Mon, 2 Dec 2002 10:25:49 -0500 (EST) Delivered-To: idr@merit.edu Received: from presque.djinesys.com (dns.nexthop.com [64.211.218.216]) by segue.merit.edu (Postfix) with ESMTP id 477825DDE0 for <idr@merit.edu>; Mon, 2 Dec 2002 10:25:49 -0500 (EST) Received: (from root@localhost) by presque.djinesys.com (8.11.3/8.11.1) id gB2FPj473646; Mon, 2 Dec 2002 10:25:45 -0500 (EST) (envelope-from jhaas@jhaas.nexthop.com) Received: from jhaas.nexthop.com (jhaas.nexthop.com [64.211.218.31]) by presque.djinesys.com (8.11.3/8.11.1) with ESMTP id gB2FPfC73634; Mon, 2 Dec 2002 10:25:41 -0500 (EST) (envelope-from jhaas@jhaas.nexthop.com) Received: (from jhaas@localhost) by jhaas.nexthop.com (8.11.3nb1/8.11.3) id gB2FPeO14605; Mon, 2 Dec 2002 10:25:40 -0500 (EST) Date: Mon, 2 Dec 2002 10:25:40 -0500 From: Jeffrey Haas <jhaas@nexthop.com> To: Manav <manav@samsung.com> Cc: idr@merit.edu Subject: Re: -18 last call comments Message-ID: <20021202102540.A14346@nexthop.com> References: <00bf01c29a46$503abc50$8946d5a5@samsungy77z6fd> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <00bf01c29a46$503abc50$8946d5a5@samsungy77z6fd>; from manav@samsung.com on Mon, Dec 02, 2002 at 01:03:50PM -0800 X-Virus-Scanned: by AMaViS perl-11 Sender: owner-idr@merit.edu Precedence: bulk Manav, On Mon, Dec 02, 2002 at 01:03:50PM -0800, Manav wrote: > I understand that there can be situations when you receive your own AS > number in the UPDATE message and the best course of action is to simply > ignore it and proceed with the next set. In the case of receiving your own AS number, this is fairly typical. You may wish to search the archives for the thread on outbound as-loop detection. > But I personally feel uncomfortable > at simply ignoring UPDATE messages without doing anything else! This is a case of doing the thing that is least annoying to the user. *Sure* you could log the messages, but in the above case, you'd end up saturating your logs with useless messages. > I agree, > that tearing down a session based solely on one UPDATE message is rather > unfair but isn't there something which we should do to inform our remote > peer that it is sending us UPDATEs which have our AS nos in the AS Path. But this is normal behavior. If I have some route which I have selected as active, I will propagate this route to you and prepend my AS number to the AS PATH. Nothing in the current protocol specification says that I should check the outgoing AS Path to see if *your* AS number is in it before I send it to you. Note that actually *doing* the outbound loop detection may speed convergence and is completely interoperable with existing implementations. > I remember there was a discussion going on some time back in this list over > such non destructive messages (I think it was INFORM messages) and it > somehow never clicked! Are we sure that we *don't* require any such > mechanisms? In this particular case, definitely not. > Regards, > Manav -- Jeff Haas NextHop Technologies Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id JAA16224 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 09:55:41 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id B538A91255; Mon, 2 Dec 2002 09:55:14 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 84E8691256; Mon, 2 Dec 2002 09:55:14 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 25A4791255 for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 09:55:13 -0500 (EST) Received: by segue.merit.edu (Postfix) id 0B6695E002; Mon, 2 Dec 2002 09:55:13 -0500 (EST) Delivered-To: idr@merit.edu Received: from merlot.juniper.net (natint.juniper.net [207.17.136.129]) by segue.merit.edu (Postfix) with ESMTP id 95DDB5DFCC for <idr@merit.edu>; Mon, 2 Dec 2002 09:55:12 -0500 (EST) Received: from juniper.net (garnet.juniper.net [172.17.28.17]) by merlot.juniper.net (8.11.3/8.11.3) with ESMTP id gB2EsYS39593; Mon, 2 Dec 2002 06:54:34 -0800 (PST) (envelope-from yakov@juniper.net) Message-Id: <200212021454.gB2EsYS39593@merlot.juniper.net> To: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com> Cc: Alex Zinin <zinin@psg.com>, idr@merit.edu Subject: Re: proxy: more comments on draft -18 In-Reply-To: Your message of "Mon, 02 Dec 2002 17:23:28 +0530." <55E277B99171E041ABF5F4B1C6DDCA06DB41CC@haritha.hclt.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <58934.1038840874.1@juniper.net> Date: Mon, 02 Dec 2002 06:54:34 -0800 From: Yakov Rekhter <yakov@juniper.net> Sender: owner-idr@merit.edu Precedence: bulk Siva, > My comments are inline. > > > > > > > Section 1, page 5: IGP definition - it's not clear from this > > > definition whether IBGP would be considered an IGP? > > > > any suggestion on how to improve the definition to clarify > > this issue would be appreciated. > > IBGP is not a protocol, so is such a clarification really required ? Just to add, a reader possessing a thorough understanding of foundations of this work, including IP routing, and the referenced documents wouldn't consider IBGP as an IGP. > If it will make things clearer for readers, however, how about a simple > clarification along the lines of: > > The usage of BGP to distribute routing information within an AS does not > fall under this definition of an IGP. > > > > > > Section 3, page 7, para 4: Does RFC 1772 still represent > > the *planned* > > > use of BGP? Or the actual use? Or something different from actual > > > use? > > > > Perhaps we should just take out references to 1772. > > My opinion is this should stay, because it does act as pointer to useful > information on this topic. It will stay, but just in the reference section. > > > Section 3, page 8, para 3 - "The hosts executing..." This paragraph > > > seems obsolete. > > > > I'll take it out. > > Don't some route optimization vendors sell boxes that operate on this > basis ? > > > > Section 4.3, Page 14/15: all multi-byte length fields - > > network byte order > > Do you think it is useful to state once at the beginning of the document > that all fields are considered to be in network order ? I'll do this for the next revision. > > > Section 4.4, Page 20: - "BGP does not use any TCP-based keep-alive > > > mechanism to determine if peers are reachable". Is it worth noting > > > that TCP may still timeout the connection even if TCP keepalives are > > > turned off? > > > > the text is fine as it is. > > I concur. > > > > > > Section 9.1.2.2, page 63, paragraph starting "If a MULTI_EXIT_DISC > > > attribute is removed before..." The first sentence is pretty nearly > > > incomprehensible. > > > > here is my attempt to clarify this: > > > > If a MULTI_EXIT_DISC attribute is removed before re-advertising a > > route into IBGP, then (prior to the removal) the MULTI_EXIT_DISC > > attribute may only be considered in the comparison of EBGP learned > > routes; the attribute is then removed, and then the remaining EBGP > > learned routes may be compared to the remaining IBGP learned routes, > > without considering the MULTI_EXIT_DISC attribute for those EBGP > > learned routes whose MULTI_EXIT_DISC attribute will be removed > > before advertising these routes to IBGP. > > > > Any further suggestions on how to improve this would be appreciated. > > How about this: > > % If a MULTI_EXIT_DISC attribute is removed before re-advertising a > % route into IBGP, then comparison based on the MULT_EXIT_DISC > % attribute may (MUST?) be performed only among the EBGP learned routes. > % This comparison MUST be performed before the removal of the > % MULTI_EXIT_DISC attribute. The MULT_EXIT_DISC attribute must then > % be removed from those EBGP routes where such removal is required and > % which are still eligible. This is followed by comparison with IBGP > % learned routes. > > I think this reflects our objectives, which is: > > a) If MED is to be removed, compare EBGP routes based on the MED > > b) Then remove the MED > > c) Then do comparison with IBGP routes If nobody would propose anything better, then I'll use your text in the next revision. > > > Section 9.3 - shouldn't this subsection be moved to the > > discussion of > > > Phase 1 or Phase 2 of the decision process? Or at least move it > > > before Section 9.2. > > > > I think it is fine where it is now. > > I concur. Thanks for your comments. Yakov. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id HAA10739 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 07:48:43 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 563089122E; Mon, 2 Dec 2002 07:48:24 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 2C3E091251; Mon, 2 Dec 2002 07:48:24 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 2F1399122E for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 07:47:56 -0500 (EST) Received: by segue.merit.edu (Postfix) id 144E55DFF0; Mon, 2 Dec 2002 07:47:56 -0500 (EST) Delivered-To: idr@merit.edu Received: from mail.alcatel.be (alc250.alcatel.be [195.207.101.250]) by segue.merit.edu (Postfix) with ESMTP id 4F99A5DE2A for <idr@merit.edu>; Mon, 2 Dec 2002 07:47:54 -0500 (EST) Received: from Bemail06.net.alcatel.be (relay3 [127.0.0.1]) by mail.alcatel.be (8.11.0/8.11.4) with ESMTP id gB2ClhD12588; Mon, 2 Dec 2002 13:47:43 +0100 Received: from alcatel.be ([138.203.191.116]) by Bemail06.net.alcatel.be (Lotus Domino Release 5.0.9a) with ESMTP id 2002120213474176:3576 ; Mon, 2 Dec 2002 13:47:41 +0100 Message-ID: <3DEB5660.ADFBAC31@alcatel.be> Date: Mon, 02 Dec 2002 13:47:29 +0100 From: hans.de_vleeschouwer@alcatel.be X-Mailer: Mozilla 4.72 [en] (WinNT; I) X-Accept-Language: en MIME-Version: 1.0 To: Manav <manav@samsung.com> Cc: "John G. Scudder" <jgs@cisco.com>, idr@merit.edu Subject: Re: -18 last call comments References: <00bf01c29a46$503abc50$8946d5a5@samsungy77z6fd> X-MIMETrack: Itemize by SMTP Server on BEMAIL06/BE/ALCATEL(Release 5.0.9a |January 7, 2002) at 12/02/2002 13:47:41, Serialize by Router on BEMAIL06/BE/ALCATEL(Release 5.0.9a |January 7, 2002) at 12/02/2002 13:47:43, Serialize complete at 12/02/2002 13:47:43 Content-Transfer-Encoding: 7bit Content-Type: text/plain; charset=us-ascii Sender: owner-idr@merit.edu Precedence: bulk Manav wrote: > > Right. In fact this is nothing new. The existence of the "AS > > Routing Loop" error incorrectly implied that such an UPDATE was a > > protocol violation, which it is not. > > I understand that there can be situations when you receive your own AS > number in the UPDATE message and the best course of action is to simply > ignore it and proceed with the next set. But I personally feel uncomfortable > at simply ignoring UPDATE messages without doing anything else! In case of VPNs, it seems that loop prevention based on AS-path is sometimes disabled and replaced by loop prevention based on 'site of origin' to make sure that routes learned from a site (via eBGP) are not readvertised to that site. (example of hub and spoke VPN architecture). In this case it seems that ignoring the UPDATE is a fair thing to do. regards, Hans De Vleeschouwer Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id GAA08519 for <idr-archive@nic.merit.edu>; Mon, 2 Dec 2002 06:51:45 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 88BAF9122C; Mon, 2 Dec 2002 06:51:17 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 4E72691250; Mon, 2 Dec 2002 06:51:17 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id C6BB39122C for <idr@trapdoor.merit.edu>; Mon, 2 Dec 2002 06:51:15 -0500 (EST) Received: by segue.merit.edu (Postfix) id AD80A5DFEB; Mon, 2 Dec 2002 06:51:15 -0500 (EST) Delivered-To: idr@merit.edu Received: from ganesh.ctd.hctech.com (unknown [202.54.64.2]) by segue.merit.edu (Postfix) with ESMTP id 223685DD8C for <idr@merit.edu>; Mon, 2 Dec 2002 06:51:14 -0500 (EST) Received: by GANESH with Internet Mail Service (5.5.2653.19) id <YCXKTHJT>; Mon, 2 Dec 2002 17:29:25 +0530 Message-ID: <55E277B99171E041ABF5F4B1C6DDCA06DB41CC@haritha.hclt.com> From: "Sivananda Ramnath - CTD, Chennai." <siva@ctd.hcltech.com> To: Yakov Rekhter <yakov@juniper.net>, Alex Zinin <zinin@psg.com> Cc: idr@merit.edu Subject: RE: proxy: more comments on draft -18 Date: Mon, 2 Dec 2002 17:23:28 +0530 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="iso-8859-1" Sender: owner-idr@merit.edu Precedence: bulk Hello, My comments are inline. > > > > Section 1, page 5: IGP definition - it's not clear from this > > definition whether IBGP would be considered an IGP? > > any suggestion on how to improve the definition to clarify > this issue would be appreciated. IBGP is not a protocol, so is such a clarification really required ? If it will make things clearer for readers, however, how about a simple clarification along the lines of: The usage of BGP to distribute routing information within an AS does not fall under this definition of an IGP. > > > Section 3, page 7, para 4: Does RFC 1772 still represent > the *planned* > > use of BGP? Or the actual use? Or something different from actual > > use? > > Perhaps we should just take out references to 1772. My opinion is this should stay, because it does act as pointer to useful information on this topic. > > > Section 3, page 8, para 3 - "The hosts executing..." This paragraph > > seems obsolete. > > I'll take it out. Don't some route optimization vendors sell boxes that operate on this basis ? > > > Section 4.3, Page 14/15: all multi-byte length fields - > network byte order Do you think it is useful to state once at the beginning of the document that all fields are considered to be in network order ? > > > > > Section 4.4, Page 20: - "BGP does not use any TCP-based keep-alive > > mechanism to determine if peers are reachable". Is it worth noting > > that TCP may still timeout the connection even if TCP keepalives are > > turned off? > > the text is fine as it is. I concur. > > > Section 9.1.2.2, page 63, paragraph starting "If a MULTI_EXIT_DISC > > attribute is removed before..." The first sentence is pretty nearly > > incomprehensible. > > here is my attempt to clarify this: > > If a MULTI_EXIT_DISC attribute is removed before re-advertising a > route into IBGP, then (prior to the removal) the MULTI_EXIT_DISC > attribute may only be considered in the comparison of EBGP learned > routes; the attribute is then removed, and then the remaining EBGP > learned routes may be compared to the remaining IBGP learned routes, > without considering the MULTI_EXIT_DISC attribute for those EBGP > learned routes whose MULTI_EXIT_DISC attribute will be removed > before advertising these routes to IBGP. > > Any further suggestions on how to improve this would be appreciated. How about this: % If a MULTI_EXIT_DISC attribute is removed before re-advertising a % route into IBGP, then comparison based on the MULT_EXIT_DISC % attribute may (MUST?) be performed only among the EBGP learned routes. % This comparison MUST be performed before the removal of the % MULTI_EXIT_DISC attribute. The MULT_EXIT_DISC attribute must then % be removed from those EBGP routes where such removal is required and % which are still eligible. This is followed by comparison with IBGP % learned routes. I think this reflects our objectives, which is: a) If MED is to be removed, compare EBGP routes based on the MED b) Then remove the MED c) Then do comparison with IBGP routes > > > Section 9.3 - shouldn't this subsection be moved to the > discussion of > > Phase 1 or Phase 2 of the decision process? Or at least move it > > before Section 9.2. > > I think it is fine where it is now. I concur. Siva (siva@ctd.hcltech.com) Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id XAA15253 for <idr-archive@nic.merit.edu>; Sun, 1 Dec 2002 23:18:36 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id CDD929124E; Sun, 1 Dec 2002 23:18:17 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 997F39124F; Sun, 1 Dec 2002 23:18:17 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 4869E9124E for <idr@trapdoor.merit.edu>; Sun, 1 Dec 2002 23:18:16 -0500 (EST) Received: by segue.merit.edu (Postfix) id 28DAE5DECE; Sun, 1 Dec 2002 23:18:16 -0500 (EST) Delivered-To: idr@merit.edu Received: from realname (unknown [203.254.224.25]) by segue.merit.edu (Postfix) with ESMTP id C90755DE74 for <idr@merit.edu>; Sun, 1 Dec 2002 23:18:15 -0500 (EST) Received: from custom-daemon.mailout2.samsung.com by mailout2.samsung.com (iPlanet Messaging Server 5.2 HotFix 1.05 (built Nov 6 2002)) id <0H6H006055JNJD@mailout2.samsung.com> for idr@merit.edu; Mon, 02 Dec 2002 13:23:47 +0900 (KST) Received: from ep_mmp1 (localhost [127.0.0.1]) by mailout2.samsung.com (iPlanet Messaging Server 5.2 HotFix 1.05 (built Nov 6 2002)) with ESMTP id <0H6H00G835JNQS@mailout2.samsung.com> for idr@merit.edu; Mon, 02 Dec 2002 13:23:47 +0900 (KST) Received: from samsungy77z6fd ([165.213.70.137]) by mmp1.samsung.com (iPlanet Messaging Server 5.2 HotFix 1.05 (built Nov 6 2002)) with ESMTPA id <0H6H00I835L8WB@mmp1.samsung.com> for idr@merit.edu; Mon, 02 Dec 2002 13:24:44 +0900 (KST) Date: Mon, 02 Dec 2002 13:03:50 -0800 From: Manav <manav@samsung.com> Subject: Re: -18 last call comments To: "John G. Scudder" <jgs@cisco.com>, idr@merit.edu Message-id: <00bf01c29a46$503abc50$8946d5a5@samsungy77z6fd> MIME-version: 1.0 X-MIMEOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 X-Mailer: Microsoft Outlook Express 5.00.2919.6700 Content-type: text/plain; charset=iso-8859-1 Content-transfer-encoding: 7BIT X-Priority: 3 X-MSMail-priority: Normal Sender: owner-idr@merit.edu Precedence: bulk > Right. In fact this is nothing new. The existence of the "AS > Routing Loop" error incorrectly implied that such an UPDATE was a > protocol violation, which it is not. I understand that there can be situations when you receive your own AS number in the UPDATE message and the best course of action is to simply ignore it and proceed with the next set. But I personally feel uncomfortable at simply ignoring UPDATE messages without doing anything else! I agree, that tearing down a session based solely on one UPDATE message is rather unfair but isn't there something which we should do to inform our remote peer that it is sending us UPDATEs which have our AS nos in the AS Path. Some sort of no destructive CEASE message with an appropriate sub-code. I am sure there can be plenty of other uses too! I remember there was a discussion going on some time back in this list over such non destructive messages (I think it was INFORM messages) and it somehow never clicked! Are we sure that we *don't* require any such mechanisms? Just a thought. Regards, Manav Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id WAA13514 for <idr-archive@nic.merit.edu>; Sun, 1 Dec 2002 22:47:18 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 1EC529124D; Sun, 1 Dec 2002 22:46:47 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 9C60B9124E; Sun, 1 Dec 2002 22:46:42 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 84E169124D for <idr@trapdoor.merit.edu>; Sun, 1 Dec 2002 22:46:28 -0500 (EST) Received: by segue.merit.edu (Postfix) id 588625DFC4; Sun, 1 Dec 2002 22:46:28 -0500 (EST) Delivered-To: idr@merit.edu Received: from cisco.com (router.cisco.com [171.69.182.20]) by segue.merit.edu (Postfix) with ESMTP id 8D63F5E085 for <idr@merit.edu>; Sun, 1 Dec 2002 22:46:27 -0500 (EST) Received: from [192.168.42.10] (ssh-rtp-1.cisco.com [161.44.11.166]) by cisco.com (8.8.8/2.6/Cisco List Logging/8.8.8) with ESMTP id WAA26120; Sun, 1 Dec 2002 22:46:18 -0500 (EST) Mime-Version: 1.0 X-Sender: jgs@router Message-Id: <p05200f0bba10872ad660@[192.168.42.10]> In-Reply-To: <20021202020508.52732.qmail@web20302.mail.yahoo.com> References: <20021202020508.52732.qmail@web20302.mail.yahoo.com> Date: Sun, 1 Dec 2002 22:46:21 -0500 To: Mareline Sheldon <marelines@yahoo.com> From: "John G. Scudder" <jgs@cisco.com> Subject: Re: -18 last call comments Cc: Yakov Rekhter <yakov@juniper.net>, Jeffrey Haas <jhaas@nexthop.com>, idr@merit.edu Content-Type: text/plain; charset="us-ascii" ; format="flowed" Sender: owner-idr@merit.edu Precedence: bulk At 6:05 PM -0800 12/1/02, Mareline Sheldon wrote: >Hi, >> From Appendix A: >> >> UPDATE Message Error subcode 7 (AS Routing Loop) has been deprecated. > >Can you please tell me the reason why this subcode has been >deprecated? As i understand it means that if i now recieve this >UPDATE which carries the same AS as mine then i need not close the >session. I just need to ignore this particular UPDATE. Right. In fact this is nothing new. The existence of the "AS Routing Loop" error incorrectly implied that such an UPDATE was a protocol violation, which it is not. >Incidentally, i was doing some testing yesterday and was confused >with one major vendor's BGP implementation. It is accepting UPDATEs >which carry its own AS number in the AS-SET and advertises them to >its EBGP peers. It refuses these UPDATEs if it finds its own AS >number in the AS SEQ and does not advertise them to its EBGP >neighbors. > >AFAIK, a BGP implementation should check for its own AS nos. in both >the AS-SET and AS-SEQ. Am i missing something here? What you describe sounds like a bug in the implementation to me. --John Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id VAA07615 for <idr-archive@nic.merit.edu>; Sun, 1 Dec 2002 21:05:38 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 10A5391249; Sun, 1 Dec 2002 21:05:12 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id CE59B9124B; Sun, 1 Dec 2002 21:05:11 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id AE0B691249 for <idr@trapdoor.merit.edu>; Sun, 1 Dec 2002 21:05:10 -0500 (EST) Received: by segue.merit.edu (Postfix) id 885325DDB3; Sun, 1 Dec 2002 21:05:10 -0500 (EST) Delivered-To: idr@merit.edu Received: from web20302.mail.yahoo.com (web20302.mail.yahoo.com [216.136.226.83]) by segue.merit.edu (Postfix) with SMTP id D30CF5DDF9 for <idr@merit.edu>; Sun, 1 Dec 2002 21:05:09 -0500 (EST) Message-ID: <20021202020508.52732.qmail@web20302.mail.yahoo.com> Received: from [165.213.1.1] by web20302.mail.yahoo.com via HTTP; Sun, 01 Dec 2002 18:05:08 PST Date: Sun, 1 Dec 2002 18:05:08 -0800 (PST) From: Mareline Sheldon <marelines@yahoo.com> Subject: Re: -18 last call comments To: Yakov Rekhter <yakov@juniper.net>, Jeffrey Haas <jhaas@nexthop.com> Cc: idr@merit.edu In-Reply-To: <200212010046.gB10kvS75662@merlot.juniper.net> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Sender: owner-idr@merit.edu Precedence: bulk Hi, > From Appendix A: > > UPDATE Message Error subcode 7 (AS Routing Loop) has been deprecated. Can you please tell me the reason why this subcode has been deprecated? As i understand it means that if i now recieve this UPDATE which carries the same AS as mine then i need not close the session. I just need to ignore this particular UPDATE. Incidentally, i was doing some testing yesterday and was confused with one major vendor's BGP implementation. It is accepting UPDATEs which carry its own AS number in the AS-SET and advertises them to its EBGP peers. It refuses these UPDATEs if it finds its own AS number in the AS SEQ and does not advertise them to its EBGP neighbors. AFAIK, a BGP implementation should check for its own AS nos. in both the AS-SET and AS-SEQ. Am i missing something here? with regards, mareline s. __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id MAA08806 for <idr-archive@nic.merit.edu>; Sun, 1 Dec 2002 12:17:12 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 1028991247; Sun, 1 Dec 2002 12:16:54 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 7278891249; Sun, 1 Dec 2002 12:16:53 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 04BDD91247 for <idr@trapdoor.merit.edu>; Sun, 1 Dec 2002 12:16:49 -0500 (EST) Received: by segue.merit.edu (Postfix) id D2C225E054; Sun, 1 Dec 2002 12:16:49 -0500 (EST) Delivered-To: idr@merit.edu Received: from merlot.juniper.net (natint.juniper.net [207.17.136.129]) by segue.merit.edu (Postfix) with ESMTP id 5B3265E02C for <idr@merit.edu>; Sun, 1 Dec 2002 12:16:49 -0500 (EST) Received: from juniper.net (garnet.juniper.net [172.17.28.17]) by merlot.juniper.net (8.11.3/8.11.3) with ESMTP id gB1HGIS00373; Sun, 1 Dec 2002 09:16:18 -0800 (PST) (envelope-from yakov@juniper.net) Message-Id: <200212011716.gB1HGIS00373@merlot.juniper.net> To: Alex Zinin <zinin@psg.com> Cc: idr@merit.edu Subject: Re: proxy: more comments on draft -18 In-Reply-To: Your message of "Sat, 30 Nov 2002 21:52:34 PST." <41117645615.20021130215234@psg.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <80846.1038762978.1@juniper.net> Date: Sun, 01 Dec 2002 09:16:18 -0800 From: Yakov Rekhter <yakov@juniper.net> Sender: owner-idr@merit.edu Precedence: bulk Alex, > Comments from another rtg-dir reviewer. > May overlap with those already discussed-- > didn't check latest threads. > > Thanks. > > -- > Alex > > Almost all of these are simple clarifications. > > Section 1, page 5: IGP definition - it's not clear from this > definition whether IBGP would be considered an IGP? any suggestion on how to improve the definition to clarify this issue would be appreciated. > Section 3, page 7, para 4: Does RFC 1772 still represent the *planned* > use of BGP? Or the actual use? Or something different from actual > use? Perhaps we should just take out references to 1772. > Section 3, page 8, para 3 - "The hosts executing..." This paragraph > seems obsolete. I'll take it out. > Section 4.1, page 11 - Length is in network byte order. all the encodings are in network byte order. This applies not just to the BGP spec, but to other protocols as well. > Section 4.2, page 12 - "My autonomous system" and "Hold time" are in > network byte order. > > Section 4.2, page 12 - Hold Time - what does a value of zero indicate? if you read section 4.4 then you'll find that: If the negotiated Hold Time interval is zero, then periodic KEEPALIVE messages MUST NOT be sent. > Section 4.2, page 13 - BGP Identifier - network byte order? > "IP address" -> "IPv4 address" I'll put at the beginning a sentence saying that in the context of this document the term "IP address" means an IP Version 4 address. > > This section should probably comment on the globally > routable/uniqueness requirements of BGP ID. > > Section 4.3, Page 14, para1, sentence 2 - "path attribute" -> "path > attributes" fixed. > Section 4.3, Page 14/15: all multi-byte length fields - network byte order > > Section 4.3, Page 16 - two byte extended length is in network byte order. > > Section 4.3,Page 17, NEXT_HOP: "IP address" -> "IPv4 address" > Specify that this is 4 octets. > Reference here to multi-protocol extensions for IPv6 nexthop? no. > RFC 2283 is unclear whether NEXT_HOP should always be included > when using multiprotocol extensions. Clarify this here? It is already clarified in 2283bis. > Section 4.3, Page 17/18 - MED and LocalPref: > "non-negative" -> "unsigned" for consistency with elsewhere. > (non-negative might imply values > 2^31 cannot be used). fixed. > Section 4.3, Page 18 - AGGREGATOR: "IP address" -> "IPv4 address". > > Section 4.3, Page 19 - Prefix: "IP address" -> "IPv4 address" > Prefix: "enough trailing bits to" -> "the minimum number of > trailing bits needed to" fixed. > > Section 4.4, Page 20: - "BGP does not use any TCP-based keep-alive > mechanism to determine if peers are reachable". Is it worth noting > that TCP may still timeout the connection even if TCP keepalives are > turned off? the text is fine as it is. > Section 4.4, Page 20: > KEEPALIVE message consists" -> "A KEEPALIVE message consists" fixed. > Section 5, Page 23: "The same attribute can not appear more than once > with the Path Attributes field...". Does this mean the same attribute > type, or the same attribute type and value? the former (the same attribute type). > Section 5.1 "The usage of each BGP path attributes .." -> attribute fixed. > Section 5.1.3 "IP address" -> "IPv4 address" > > "A BGP speaker must never advertise an address of a peer to that peer > as a NEXT_HOP, for a route that the speaker is originating." > suggest replace this text with: > "A route originated by a BGP speaker must never be advertised to a > peer using an address of that peer as NEXT_HOP" fixed. > > Section 5.1.4: "A BGP speaker MUST IMPLEMENT a mechanism ... which > allows the MULTI_EXIT_DISC to be removed from a route." Might want to > say that this is dangerous unless you received the route from an EBGP > peer? I think we should keep the text as is. > Section 5.1.5: "If it [LOCAL_PREF] is contained in an UPDATE message > that is received from an external peer, then this attribute MUST be > ignored by the receiving speaker, except for the case of BGP > Confederations [RF3065]." > - "ignored" might be taken to mean that you don't process it for > decision, but that you propagate it to internal peers. I might > write "silently removed" or something similar. I think the text is ok as is. > Section 5.1.5, para 2. "set of AS" -> "set of ASs" fixed. > > Section 5.1.7: "IP address" -> "IPv4 address" > > Secton 6.2: note that Data field is in network byte order. > > Section 6.2, para 5: "IP host address" -> "IPv4 host address" > > Section 6.3, wrt NEXT_HOP: "IP address" -> "IPv4 address" > > Section 6.3: wrt NEXT_HOP semantic correctness: should we check that a > NEXT_HOP is not a multicast or broadcast address? I'll add to the definition of NEXT_HOP that it is a unicast address. > Section 6.3, page 32, para 7: "peer than sent" -> "peer that sent" fixed. > Section 6.3: "if any attribute appears more than once" - does this > mean the same attribute type, or the same attribute type and value? the former. > Section 6.8 "Comparing BGP identifiers is done by treating them as > (4-octet-long) unsigned integers". Need to convert to host byte order > before comparing. fixed. > Section 6.8, item 2: "closes BGP connection" -> "closes the BGP connection" > "accepts BGP connection" -> "accepts the BGP connection". fixed. > Section 9.1.2.2: item (c): in the explanation of neighborAS(n), it is > unclear for IBGP connections how to determine "the neighbor AS from > which the other IBGP speaker learned the route". If this is really > the leftmost entry in the AS path (or the local AS if the path is > empty), the spec should explicitly say so. fixed. > Section 9.1.2.2, page 63, paragraph starting "If a MULTI_EXIT_DISC > attribute is removed before..." The first sentence is pretty nearly > incomprehensible. here is my attempt to clarify this: If a MULTI_EXIT_DISC attribute is removed before re-advertising a route into IBGP, then (prior to the removal) the MULTI_EXIT_DISC attribute may only be considered in the comparison of EBGP learned routes; the attribute is then removed, and then the remaining EBGP learned routes may be compared to the remaining IBGP learned routes, without considering the MULTI_EXIT_DISC attribute for those EBGP learned routes whose MULTI_EXIT_DISC attribute will be removed before advertising these routes to IBGP. Any further suggestions on how to improve this would be appreciated. > Section 9.1.2.2 (d) > "d) If at least one of the candidate routes was received from an > external peer in a neighboring autonomous system, remove from con- > sideration all routes which were received from internal peers." > For consistency with (c) and clarity, this might be reworded: > "d) If any of the candidate routes was learned via EBGP, remove > from consideration all routes which were learned by IBGP." fixed. > > Section 9.1.2.2 (e) > "cost (n) is better than cost (m)" > Given the definition of cost, it might be clearer to say > "cost (n) is lower than cost (m)" fixed. > Section 9.1.2.2 (g) > "neighbor address" has not been defined. I'll replace "neighbor address" with "peer address". > Section 9.2.2.2, Page 70 (AGGREGATOR) - "All AGGREGATOR attributes of > all routes to be aggregated should be ignored." > > Perhaps "ignored" is ambiguous here, and it's not clear whether > should is a SHOULD. Suggest: > > "Any AGGREGATOR attributes from the routes to be aggregated MUST > NOT be included in the aggregated route." fixed. > Section 9.3 - shouldn't this subsection be moved to the discussion of > Phase 1 or Phase 2 of the decision process? Or at least move it > before Section 9.2. I think it is fine where it is now. > > Appendix E, para 2: IP precedence has been deprecated. Delete this > paragraph, or replace with appropriate diffserv codepoint. deleted. > Security Considerations: > "BGP supports the ability to authenticate BGP messages by using BGP > authentication." > This sentence should be removed, and the Authentication Information > parameter has been deprecated. Please see the recent e-mail exchange on the Security Considerations Yakov. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id LAA04651 for <idr-archive@nic.merit.edu>; Sun, 1 Dec 2002 11:02:07 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id 9C2FF9121B; Sun, 1 Dec 2002 11:01:53 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 27AC391247; Sun, 1 Dec 2002 11:01:44 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id DD0589121B for <idr@trapdoor.merit.edu>; Sun, 1 Dec 2002 11:01:22 -0500 (EST) Received: by segue.merit.edu (Postfix) id B25965DFA7; Sun, 1 Dec 2002 11:01:22 -0500 (EST) Delivered-To: idr@merit.edu Received: from merlot.juniper.net (natint.juniper.net [207.17.136.129]) by segue.merit.edu (Postfix) with ESMTP id 502645DE07 for <idr@merit.edu>; Sun, 1 Dec 2002 11:01:22 -0500 (EST) Received: from juniper.net (garnet.juniper.net [172.17.28.17]) by merlot.juniper.net (8.11.3/8.11.3) with ESMTP id gB1G0pS98632; Sun, 1 Dec 2002 08:00:51 -0800 (PST) (envelope-from yakov@juniper.net) Message-Id: <200212011600.gB1G0pS98632@merlot.juniper.net> To: Alex Zinin <zinin@psg.com> Cc: idr@merit.edu Subject: Re: proxy: comments on draft -18 In-Reply-To: Your message of "Sat, 30 Nov 2002 21:49:14 PST." <51117445637.20021130214914@psg.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-ID: <65472.1038758451.1@juniper.net> Date: Sun, 01 Dec 2002 08:00:51 -0800 From: Yakov Rekhter <yakov@juniper.net> Sender: owner-idr@merit.edu Precedence: bulk Alex, > Sorry for late posting--been traveling. > Below are comments from a rtg-dir reviewer. > Thanks. > > -- > Alex > > A few comments > > (i). I agree with the concern voiced on the IDR list > by Tom Petch (and possibly others), namely, that > it appears that the FSM specification is > inconsistent when describing the action that > places a value in a MIB object when a > NOTIFICATION is sent. See Jeff's reply to Tom on this topic. > (ii). It might be useful for section 3 to describe the > relationship between the various (conceptual) > RIBs (Adj-RIB-In, Adj-RIB-Out, and Loc-RIB) and > what we are currently calling a FIB. This, > however, is not critical to the protocol > specification. and therefore the text will stay as is. > (iii). Page 23, 3rd paragraph says "The sender of an > UPDATE message should order ..." > > Is this "should" a "SHOULD" (in the RFC 2119 > sense)? Likewise section 5.1.2 (AS_PATH), section > a): "the advertising speaker shall not > modifiy..." > > Actually, there seems to be a lot of use of the > terms "shall" and "shall not" which are not > capitalized, so it is unclear what the exact > meaning is. "may" is also used in the same way. Will fix in the next revision. Yakov. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id AAA04887 for <idr-archive@nic.merit.edu>; Sun, 1 Dec 2002 00:53:40 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id A444D91245; Sun, 1 Dec 2002 00:53:04 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 6FD5591246; Sun, 1 Dec 2002 00:53:04 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 067CD91245 for <idr@trapdoor.merit.edu>; Sun, 1 Dec 2002 00:53:02 -0500 (EST) Received: by segue.merit.edu (Postfix) id D9B195DE96; Sun, 1 Dec 2002 00:53:02 -0500 (EST) Delivered-To: idr@merit.edu Received: from psg.com (psg.com [147.28.0.62]) by segue.merit.edu (Postfix) with ESMTP id 5BE025DDA6 for <idr@merit.edu>; Sun, 1 Dec 2002 00:53:02 -0500 (EST) Received: from psg.com ([147.28.0.62] helo=127.0.0.1 ident=zinin) by psg.com with esmtp (Exim 3.36 #2) id 18IN2L-000Im5-00 for idr@merit.edu; Sat, 30 Nov 2002 21:53:01 -0800 Date: Sat, 30 Nov 2002 21:52:34 -0800 From: Alex Zinin <zinin@psg.com> X-Mailer: The Bat! (v1.51) Personal Reply-To: Alex Zinin <zinin@psg.com> X-Priority: 3 (Normal) Message-ID: <41117645615.20021130215234@psg.com> To: idr@merit.edu Subject: proxy: more comments on draft -18 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-idr@merit.edu Precedence: bulk Comments from another rtg-dir reviewer. May overlap with those already discussed-- didn't check latest threads. Thanks. -- Alex Almost all of these are simple clarifications. Section 1, page 5: IGP definition - it's not clear from this definition whether IBGP would be considered an IGP? Section 3, page 7, para 4: Does RFC 1772 still represent the *planned* use of BGP? Or the actual use? Or something different from actual use? Section 3, page 8, para 3 - "The hosts executing..." This paragraph seems obsolete. Section 4.1, page 11 - Length is in network byte order. Section 4.2, page 12 - "My autonomous system" and "Hold time" are in network byte order. Section 4.2, page 12 - Hold Time - what does a value of zero indicate? Section 4.2, page 13 - BGP Identifier - network byte order? "IP address" -> "IPv4 address" This section should probably comment on the globally routable/uniqueness requirements of BGP ID. Section 4.3, Page 14, para1, sentence 2 - "path attribute" -> "path attributes" Section 4.3, Page 14/15: all multi-byte length fields - network byte order Section 4.3, Page 16 - two byte extended length is in network byte order. Section 4.3,Page 17, NEXT_HOP: "IP address" -> "IPv4 address" Specify that this is 4 octets. Reference here to multi-protocol extensions for IPv6 nexthop? RFC 2283 is unclear whether NEXT_HOP should always be included when using multiprotocol extensions. Clarify this here? Section 4.3, Page 17/18 - MED and LocalPref: "non-negative" -> "unsigned" for consistency with elsewhere. (non-negative might imply values > 2^31 cannot be used). Section 4.3, Page 18 - AGGREGATOR: "IP address" -> "IPv4 address". Section 4.3, Page 19 - Prefix: "IP address" -> "IPv4 address" Prefix: "enough trailing bits to" -> "the minimum number of trailing bits needed to" Section 4.4, Page 20: - "BGP does not use any TCP-based keep-alive mechanism to determine if peers are reachable". Is it worth noting that TCP may still timeout the connection even if TCP keepalives are turned off? " Section 4.4, Page 20: KEEPALIVE message consists" -> "A KEEPALIVE message consists" Section 5, Page 23: "The same attribute can not appear more than once with the Path Attributes field...". Does this mean the same attribute type, or the same attribute type and value? Section 5.1 "The usage of each BGP path attributes .." -> attribute Section 5.1.3 "IP address" -> "IPv4 address" "A BGP speaker must never advertise an address of a peer to that peer as a NEXT_HOP, for a route that the speaker is originating." suggest replace this text with: "A route originated by a BGP speaker must never be advertised to a peer using an address of that peer as NEXT_HOP" Section 5.1.4: "A BGP speaker MUST IMPLEMENT a mechanism ... which allows the MULTI_EXIT_DISC to be removed from a route." Might want to say that this is dangerous unless you received the route from an EBGP peer? Section 5.1.5: "If it [LOCAL_PREF] is contained in an UPDATE message that is received from an external peer, then this attribute MUST be ignored by the receiving speaker, except for the case of BGP Confederations [RF3065]." - "ignored" might be taken to mean that you don't process it for decision, but that you propagate it to internal peers. I might write "silently removed" or something similar. Section 5.1.5, para 2. "set of AS" -> "set of ASs" Section 5.1.7: "IP address" -> "IPv4 address" Secton 6.2: note that Data field is in network byte order. Section 6.2, para 5: "IP host address" -> "IPv4 host address" Section 6.3, wrt NEXT_HOP: "IP address" -> "IPv4 address" Section 6.3: wrt NEXT_HOP semantic correctness: should we check that a NEXT_HOP is not a multicast or broadcast address? Section 6.3, page 32, para 7: "peer than sent" -> "peer that sent" Section 6.3: "if any attribute appears more than once" - does this mean the same attribute type, or the same attribute type and value? Section 6.8 "Comparing BGP identifiers is done by treating them as (4-octet-long) unsigned integers". Need to convert to host byte order before comparing. Section 6.8, item 2: "closes BGP connection" -> "closes the BGP connection" "accepts BGP connection" -> "accepts the BGP connection". Section 9.1.2.2: item (c): in the explanation of neighborAS(n), it is unclear for IBGP connections how to determine "the neighbor AS from which the other IBGP speaker learned the route". If this is really the leftmost entry in the AS path (or the local AS if the path is empty), the spec should explicitly say so. Section 9.1.2.2, page 63, paragraph starting "If a MULTI_EXIT_DISC attribute is removed before..." The first sentence is pretty nearly incomprehensible. Section 9.1.2.2 (d) "d) If at least one of the candidate routes was received from an external peer in a neighboring autonomous system, remove from con- sideration all routes which were received from internal peers." For consistency with (c) and clarity, this might be reworded: "d) If any of the candidate routes was learned via EBGP, remove from consideration all routes which were learned by IBGP." Section 9.1.2.2 (e) "cost (n) is better than cost (m)" Given the definition of cost, it might be clearer to say "cost (n) is lower than cost (m)" Section 9.1.2.2 (g) "neighbor address" has not been defined. Section 9.2.2.2, Page 70 (AGGREGATOR) - "All AGGREGATOR attributes of all routes to be aggregated should be ignored." Perhaps "ignored" is ambiguous here, and it's not clear whether should is a SHOULD. Suggest: "Any AGGREGATOR attributes from the routes to be aggregated MUST NOT be included in the aggregated route." Section 9.3 - shouldn't this subsection be moved to the discussion of Phase 1 or Phase 2 of the decision process? Or at least move it before Section 9.2. Appendix E, para 2: IP precedence has been deprecated. Delete this paragraph, or replace with appropriate diffserv codepoint. Security Considerations: "BGP supports the ability to authenticate BGP messages by using BGP authentication." This sentence should be removed, and the Authentication Information parameter has been deprecated. Received: from trapdoor.merit.edu (postfix@trapdoor.merit.edu [198.108.1.26]) by nic.merit.edu (8.9.3/8.9.1) with ESMTP id AAA04697 for <idr-archive@nic.merit.edu>; Sun, 1 Dec 2002 00:50:06 -0500 (EST) Received: by trapdoor.merit.edu (Postfix) id A85CC91243; Sun, 1 Dec 2002 00:49:43 -0500 (EST) Delivered-To: idr-outgoing@trapdoor.merit.edu Received: by trapdoor.merit.edu (Postfix, from userid 56) id 6DE5B91245; Sun, 1 Dec 2002 00:49:43 -0500 (EST) Delivered-To: idr@trapdoor.merit.edu Received: from segue.merit.edu (segue.merit.edu [198.108.1.41]) by trapdoor.merit.edu (Postfix) with ESMTP id 3125891243 for <idr@trapdoor.merit.edu>; Sun, 1 Dec 2002 00:49:42 -0500 (EST) Received: by segue.merit.edu (Postfix) id 100555DE96; Sun, 1 Dec 2002 00:49:42 -0500 (EST) Delivered-To: idr@merit.edu Received: from psg.com (psg.com [147.28.0.62]) by segue.merit.edu (Postfix) with ESMTP id D80095DDA6 for <idr@merit.edu>; Sun, 1 Dec 2002 00:49:41 -0500 (EST) Received: from psg.com ([147.28.0.62] helo=127.0.0.1 ident=zinin) by psg.com with esmtp (Exim 3.36 #2) id 18IMz7-000Ign-00 for idr@merit.edu; Sat, 30 Nov 2002 21:49:41 -0800 Date: Sat, 30 Nov 2002 21:49:14 -0800 From: Alex Zinin <zinin@psg.com> X-Mailer: The Bat! (v1.51) Personal Reply-To: Alex Zinin <zinin@psg.com> X-Priority: 3 (Normal) Message-ID: <51117445637.20021130214914@psg.com> To: idr@merit.edu Subject: proxy: comments on draft -18 MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-idr@merit.edu Precedence: bulk Sorry for late posting--been traveling. Below are comments from a rtg-dir reviewer. Thanks. -- Alex A few comments (i). I agree with the concern voiced on the IDR list by Tom Petch (and possibly others), namely, that it appears that the FSM specification is inconsistent when describing the action that places a value in a MIB object when a NOTIFICATION is sent. (ii). It might be useful for section 3 to describe the relationship between the various (conceptual) RIBs (Adj-RIB-In, Adj-RIB-Out, and Loc-RIB) and what we are currently calling a FIB. This, however, is not critical to the protocol specification. (iii). Page 23, 3rd paragraph says "The sender of an UPDATE message should order ..." Is this "should" a "SHOULD" (in the RFC 2119 sense)? Likewise section 5.1.2 (AS_PATH), section a): "the advertising speaker shall not modifiy..." Actually, there seems to be a lot of use of the terms "shall" and "shall not" which are not capitalized, so it is unclear what the exact meaning is. "may" is also used in the same way.
- 13.5 in BGP Base Draft - Issue List v2.0 (Draft-1… Tom Petch
- Re: 13.5 in BGP Base Draft - Issue List v2.0 (Dra… andrewl
- Re: 13.5 in BGP Base Draft - Issue List v2.0 (Dra… Tom Petch
- Re: 13.5 in BGP Base Draft - Issue List v2.0 (Dra… andrewl
- RE: 13.5 in BGP Base Draft - Issue List v2.0 (Dra… Susan Hares
- Re: 13.5 in BGP Base Draft - Issue List v2.0 (Dra… Tom Petch