[Idr] notes from the WG meeting

Yakov Rekhter <yakov@juniper.net> Wed, 09 November 2005 16:32 UTC

Received: from localhost.cnri.reston.va.us ([] helo=megatron.ietf.org) by megatron.ietf.org with esmtp (Exim 4.32) id 1EZsrt-0006xV-Mw; Wed, 09 Nov 2005 11:32:13 -0500
Received: from odin.ietf.org ([] helo=ietf.org) by megatron.ietf.org with esmtp (Exim 4.32) id 1EZsrr-0006wD-Vz for idr@megatron.ietf.org; Wed, 09 Nov 2005 11:32:12 -0500
Received: from ietf-mx.ietf.org (ietf-mx []) by ietf.org (8.9.1a/8.9.1a) with ESMTP id LAA19996 for <idr@ietf.org>; Wed, 9 Nov 2005 11:31:43 -0500 (EST)
Received: from colo-dns-ext2.juniper.net ([]) by ietf-mx.ietf.org with esmtp (Exim 4.43) id 1EZt7u-000184-Hw for idr@ietf.org; Wed, 09 Nov 2005 11:48:47 -0500
Received: from merlot.juniper.net (merlot.juniper.net []) by colo-dns-ext2.juniper.net (8.12.3/8.12.3) with ESMTP id jA9GW2Bm056038 for <idr@ietf.org>; Wed, 9 Nov 2005 08:32:02 -0800 (PST) (envelope-from yakov@juniper.net)
Received: from juniper.net (sapphire.juniper.net []) by merlot.juniper.net (8.11.3/8.11.3) with ESMTP id jA9GW2G64941 for <idr@ietf.org>; Wed, 9 Nov 2005 08:32:02 -0800 (PST) (envelope-from yakov@juniper.net)
Message-Id: <200511091632.jA9GW2G64941@merlot.juniper.net>
To: idr@ietf.org
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <64694.1131553922.1@juniper.net>
Date: Wed, 09 Nov 2005 08:32:02 -0800
From: Yakov Rekhter <yakov@juniper.net>
X-Spam-Score: 0.0 (/)
X-Scan-Signature: 2bf730a014b318fd3efd65b39b48818c
Subject: [Idr] notes from the WG meeting
X-BeenThere: idr@ietf.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Inter-Domain Routing <idr.ietf.org>
List-Unsubscribe: <https://www1.ietf.org/mailman/listinfo/idr>, <mailto:idr-request@ietf.org?subject=unsubscribe>
List-Archive: <http://www1.ietf.org/pipermail/idr>
List-Post: <mailto:idr@ietf.org>
List-Help: <mailto:idr-request@ietf.org?subject=help>
List-Subscribe: <https://www1.ietf.org/mailman/listinfo/idr>, <mailto:idr-request@ietf.org?subject=subscribe>
Sender: idr-bounces@ietf.org
Errors-To: idr-bounces@ietf.org


Please review and comment for correctness. The deadline for
comments is Nov 21, 2005.

------- Forwarded Message

Date:    Tue, 08 Nov 2005 18:07:16 -0800
From:    Larry Blunk <lblunk@yahoo.com>
To:      skh@nexthop.com, yakov@juniper.net
Subject: notes

- --0-67043162-1131502036=:68032
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
Content-Disposition: inline

   Here are my notes.  I missed some of the exchange between Yakov and
Gargi over the BGP Connector Attribute.


Yahoo! FareChase: Search multiple travel sites in one click.
- --0-67043162-1131502036=:68032
Content-Type: text/plain; name="idr-notes.txt"
Content-Description: 4180351827-idr-notes.txt
Content-Disposition: inline; filename="idr-notes.txt"

	IDR WG Meeting Notes  -   Nov. 8, 2005   1:00 - 3:00 PM
		recorded by Larry J. Blunk  ljb@merit.edu

    Yakov - Review of document status -- refer to slides

       Enke to produce implementation reports for outbound route filtering and 
       Pekka - what is the status of the version 2 MIB?  Sue - needs two implem

   Yakov - as4octect-ext-community

        no implementations, will be published separately from ext. communities 
so as no to hold up ext. comm.
        An individual submission for now.   Should this be a WG item or should 
it die?
        Geoff Huston - don't let it die.  Want to be able to signal from 2 AS c
omm. to 4 byte AS comm.
        Consensus is to make it a WG item.  Will send to mailing list and see i
f any objections to making WG item

   Tony Li send email to list that he would like to make AS hopcount draft a WG
 item.  Pekka did not support.
       only 6-7 have read,  4 votes to make it a WG item.   Yakov - service pro
viders please read doc. and
       let people know if you want it to be WG item.

   Sue Hares - Group Cooperative Route Filtering Capability for BGP-4 - slides
     Want something like a "route-map" function for ORF.   Create ordered group
s of policies.  Apply
     group policies before non-group.  Use AND/OR bit to specify how ordered po
licy groups are applied.
     Examples provided about how this capability would be used.

     Question - Geoff - trying to construct a pseudo Boolean which is subset of
 true boolean - why not
     go other 10% to full boolean.   Feedback has been to scaleback to limited 
language.   All you need
     is to allow groups in groups.   Someone likely have a case that requires f
ull boolean logic.  Authors
     have argued about whether it is needed.  Geoff - hard to know in advance. 
  Comment -  could you
     add a NOT to make it complete?
     Enke - goal is to simplify and to improve performance.  Sue - not performa
nce issue.
     Will take back comments about missing functionality and update draft and b
ring to WG.

   David Ward - BGP for Interdomain Service Routing (aka Context AFI) - slides
     Interdomain QOS.   Led by Dave Clark at MIT.
     Yakov comment - not only BGP can't signal path characteristics - nothing c
     Want to announce a service not a packet marking - must have opaque semanti
cs for QOS bits on either side
     of AS boundary.  minimal changes to BGP - no BGP5. multisession BGP to rem
ove fate sharing.
     advertise multiple paths to same dest.  aggregate withdraw.  Context AF fo
r BGP.   QOS value 0-255.
    The ID itself is opaque and does note define a local QOS semantics.
    exchanges QOS service information - enabling service differentiation.
    Sue - this looks like old IDR QOS revisited.   Dave - what's new is always 
old.  Divergent from QOS bit.  
    Semantic free and opaque.   Service 42 is available but no meaning to seman
tics.  Sue - are you mixing
    QOS pieces?  Sue - timeliness issue.  How do you ensure announcement is the
re in time.
    Chandra Comment - could be useful to an application to group together prefi
x/service - don't need to go to IANA.
    Yakov - will you have different SAFI values for different AFI's? 
    Dave - could negotiate to RD's are all the way down to RT's, but less flexi
    Yakov - WG comment.  Bigger picture does not belong in IDR, only BGP change
s.  Yakov - should the bigger picture come
    first.   Sue - is all you are asking for is a different AFI?   Dave - Yes. 
   Add path and agg. withdraw would help, but not
    truly necessary.    Sue - AD's what is process for new AFI?   Bill - it say
s nothing, would be sent to IESG which would
    send to back IDR.  Yakov - would IESG like to see big picture?  Alex - not 
ready to answer that question, would like to see
    discussion on list.   Dave - Propose MAVs BOF to discuss QOS things not rel
ated to BGP.   Dave - this is indepedent.
    Alex - if working group discusses and believes it can be used for other tec
hnologies and there is strong support.   Yakov - there
    was discussion of tunnel setup.   Dave - I don't believe there are dependen
cies between outcome of MAVs BOF and this
    proposal.   Alex - sounds like discussion and consensus is needed.
    Bruce Davie from Cisco - feels work does stand alone if you assume there is
 a diffserv architecture.  can look at diffserv
    architecture for big picture context.   Yakov - can someone write document 
about how this fits into diffserv.    Bruce will
    find someone to do that.

   Enke Chen - BGP extended optional parameters draft - slides
     BGP capabilities in optional parameters field which is limited to 255 byte
s.   number of capabilities could exceed
     this amount.  introduce extended optional parameters.   Can carry extended
 options in open message and indicate
     whether there will be a second OPEN message with ext. optional parameters.
     Question - what if you have ORF capability?  Enke - you only do ORF after 
connection is established - will talk later.
     Sue - were state machine changes included?   John promised to do it.   Nee
ded to move forward.
     Enke - should this become a WG item?  Don't need it today, but could be im
portant tomorrow.  Only 4 or so have
     read.  Yakov wants more people to read before progressing.   Work will not
 be done if no interest.

  Chandra Appanna - BGP Aggregate Withdraw - slides
     Introduced by David.  Currently one withdraw per advertised prefix.  incom
ing withdraw processing is
     per prefix -- impacts network convergence.  send a withdraw with common at
     Ron Bonica - is there mechanism to ensure that you only withdraw routes th
at you announce?  Chandra -
     routes are only for a particular session.   Current draft is not transitiv
e, but it could be so.   Yakov -
     time-to-withdraw could be used without agg. withdraw.   Chandra - yes.
     Pekka - if we are using Secure BGP or SOBGP will it require special proces
sing?  Chandra - no, but haven't
     thought about it.   Sue - how does this relate to withdraw bags of routes,
 not sure.   Will take offline.
     Dave - with agg. withdraw you match on attributes unlike bags.   Could be 
ASPATH, Origin, etc.  Chandra -
     designed to be general.   Yakov - could mark and organize by extended comm
unities instead.   Dave - if it's
     too much flexilibility, that's okay.   Dave - WG could determine that only
 communities are needed.  Yakov - not
     clear how to propagate agg. withdraw if applied to things other than commu
nities.   Will take this offline.
     WG needs to decide if it wants to propagate agg. withdraw.

 Gargi  Nalawade - BGP Connector Attribute - slides
     Tunneling applications exist or are being proposed in BGP.   Tunnel endpoi
nt discovery.
      Most of work will be done l3vpn working group, but presenting in IDR for 
feedback here.

 Gargi  Nalawade - multicast signaling using BGP  - slides
     other work going on in other WG.  PE-PE signaling for Multicast VPNs. usin
g BGP for non-LAN based
     Multicast VPNs.  Enhancement required in BGP.  Go to L3VPN for further dis

- --0-67043162-1131502036=:68032--

------- End of Forwarded Message

Idr mailing list