The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2001-Oct> msg00346



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

ppvpn some VLAN clarifications

  • From: Anoop Ghanwani <anoop@lanterncom.com>
  • Date: Tue, 23 Oct 2001 11:07:29 -0700

 
See below for comments.
-----Original Message-----
From: Michael Wu [mailto:mwu@cisco.com]
Sent: Tuesday, October 23, 2001 9:01 AM
To: Amir Shatil; mpls@UU.NET; ppvpn@zephion.net
Subject: RE: ppvpn some VLAN clarifications

 
-----Original Message-----
From: owner-ppvpn@cthulu.dc1.dca57.zephion.net [mailto:owner-ppvpn@cthulu.dc1.dca57.zephion.net]On Behalf Of Amir Shatil
Sent: Tuesday, October 23, 2001 8:34 AM
To: mpls@uu.net; ppvpn@zephion.net
Subject: ppvpn some VLAN clarifications

Hi
few Questions about draft-martini-l2circuit-encap-mpls-03.txt: 
1. In section 4.3 it is stated (about the VLAN tag) “…and MAY be overwritten by the egress router.”
What are the cases in which this might happen? 
>> So the SP doesn't need to have same vlan id at both ends. But this will prevent STP working
   correctly unless the PE can modify the STP message accordingly. 
 
[Anoop Ghanwani]  STP doesn't use the VLAN ID. In 802.1Q, all VLANs use the same spanning
tree.  Besides I don't think the draft is trying to support bridging.  It just provides point-to-point Ethernet link.
In this case, does the VLAN tag lose its original meaning and functions only as a VC ID (more like a DLCI in F/R)? 
2. Is the reason for carrying the ethernet VLAN L2 encapsulation (section 4.3 "For an Ethernet 802.1q VLAN the entire Ethernet frame without the
   preamble or FCS is transported as a single packet") is only to preserve the original SA & DA ? 
>> 802.1p will be preserved as well.
Michael 
Thanks
Amir.