draft-ietf-l3vpn-2547bis-mcast-bgp section 13

"HEMIGE Venu" <Venu.Hemige@alcatel-lucent.com> Fri, 12 December 2008 15:54 UTC

Return-Path: <l3vpn-bounces@ietf.org>
X-Original-To: l3vpn-archive@megatron.ietf.org
Delivered-To: ietfarch-l3vpn-archive@core3.amsl.com
Received: from [127.0.0.1] (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 72AAB3A6BE4; Fri, 12 Dec 2008 07:54:05 -0800 (PST)
X-Original-To: l3vpn@core3.amsl.com
Delivered-To: l3vpn@core3.amsl.com
Received: from localhost (localhost [127.0.0.1]) by core3.amsl.com (Postfix) with ESMTP id 8824E3A6BE4 for <l3vpn@core3.amsl.com>; Fri, 12 Dec 2008 07:54:04 -0800 (PST)
X-Virus-Scanned: amavisd-new at amsl.com
X-Spam-Flag: NO
X-Spam-Score: -2.598
X-Spam-Level:
X-Spam-Status: No, score=-2.598 tagged_above=-999 required=5 tests=[BAYES_00=-2.599, HTML_MESSAGE=0.001]
Received: from mail.ietf.org ([64.170.98.32]) by localhost (core3.amsl.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id wThZ3iRmHPOc for <l3vpn@core3.amsl.com>; Fri, 12 Dec 2008 07:54:03 -0800 (PST)
Received: from audl751.usa.alcatel.com (audl751.usa.alcatel.com [143.209.238.164]) by core3.amsl.com (Postfix) with ESMTP id 7B0663A67AA for <l3vpn@ietf.org>; Fri, 12 Dec 2008 07:54:03 -0800 (PST)
Received: from usdalsbhs02.ad3.ad.alcatel.com (usdalsbhs02.usa.alcatel.com [172.22.216.13]) by audl751.usa.alcatel.com (ALCANET) with ESMTP id mBCFrp9v027270 for <l3vpn@ietf.org>; Fri, 12 Dec 2008 09:53:56 -0600
Received: from USDALSMBS03.ad3.ad.alcatel.com ([172.22.216.8]) by usdalsbhs02.ad3.ad.alcatel.com with Microsoft SMTPSVC(6.0.3790.2499); Fri, 12 Dec 2008 09:53:54 -0600
X-MimeOLE: Produced By Microsoft Exchange V6.5
Content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: multipart/alternative; boundary="----_=_NextPart_001_01C95C71.D5D17867"
Subject: draft-ietf-l3vpn-2547bis-mcast-bgp section 13
Date: Fri, 12 Dec 2008 09:53:53 -0600
Message-ID: <98E5EEEC1F4EF740912E44EA3022569A0E907B12@USDALSMBS03.ad3.ad.alcatel.com>
X-MS-Has-Attach:
X-MS-TNEF-Correlator:
Thread-Topic: draft-ietf-l3vpn-2547bis-mcast-bgp section 13
Thread-Index: AclccdUCB0Zu8P12RM+1InNxbmRFGQ==
From: HEMIGE Venu <Venu.Hemige@alcatel-lucent.com>
To: l3vpn@ietf.org
X-OriginalArrivalTime: 12 Dec 2008 15:53:54.0305 (UTC) FILETIME=[D5B21F10:01C95C71]
X-Scanned-By: MIMEDefang 2.51 on 143.209.238.34
X-BeenThere: l3vpn@ietf.org
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: <l3vpn.ietf.org>
List-Unsubscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=unsubscribe>
List-Archive: <https://www.ietf.org/mailman/private/l3vpn>
List-Post: <mailto:l3vpn@ietf.org>
List-Help: <mailto:l3vpn-request@ietf.org?subject=help>
List-Subscribe: <https://www.ietf.org/mailman/listinfo/l3vpn>, <mailto:l3vpn-request@ietf.org?subject=subscribe>
Sender: l3vpn-bounces@ietf.org
Errors-To: l3vpn-bounces@ietf.org

I think the specification in section 13 needs some enhancement.
Otherwise, there are cases where it would not work.
 
1. Section 13.2: This sections says that if an egress PE has (C-*,C-G)
states and it has "one or more" source-active routes, then it must pick
"one of them" and setup a (C-S,C-G) forwarding state to accept traffic
for that (C-S,C-G) from the originator of that source-active route. If
all PEs select a single upstream PE for an (C-S,C-G), there should be
only one source-active route available. I suppose more than one is
possible when you have a PE configured as RP and hence it sends source
active routes into the MVPN cloud. Is that correct?
 
2. Section 13.2: If the above assumption is correct, then I think it is
wrong to say that the egress PE must pick one of the source-active
routes and setup forwarding state to accept traffic from the
"originator" of that source-active route. It is possible that the
orignator may not be in the path to C-S. Even if it is, it would break
the "single upstream PE" rule. I think the RPF tunnel should be the one
from the single forwarder PE regardless of where the Soure-Active is
received from. Section 14 says this but Section 13 does not.
 
3. Section 13.2: I think it is insufficient to say that the egress PE
creates just the forwarding state for (C-S,C-G). I think the egress PE
MUST send a Source Join Route to explicitly join the (C-S, C-G) tree.
Here is why: It could be that the only reason for the Source Join Route
and hence the Source-Active in the MVPN is because C-RP sent it after
C-S sends PIM registers to C-RP. As per section 13.2.1, the upstream PE
for the (C-*,C-G) creates a (C-S,C-G,rpt) prune state and sends this
prune to the RP when it receives a Source Active. As a result, C-RP
would prune itself off the (C-S,C-G) tree and this would result in the
Source Join and Source-Active routes from getting withdrawn from the
MVPN. This would result in a repetitious cycle of Source Join,
Source-Active, S,G,Rpt prune, Withdraw Source-Join, Withdraw
Source-Active, S,G,Rpt join, ...
 
The best way to fix these issus is to require that all egress PEs that
have (C-*,C-G) states MUST send Source Join Routes when they receive a
Source-Active. 
 
Please let me know if I am missing something.
 
Thanks,
-Venu