9 messages in net.nether.puck.cisco-nsp[c-nsp] MPLS, L2TPv3 Layer 2/3 VPN Ne...
FromSent OnAttachments
Eric KaganJan 4, 2005 6:56 am 
John OsmonJan 4, 2005 11:39 am 
Nick ShahJan 4, 2005 6:41 pm 
choo...@pacific.net.sgJan 4, 2005 9:14 pm 
Nick ShahJan 4, 2005 9:26 pm 
choo...@pacific.net.sgJan 4, 2005 10:02 pm 
Jon LewisJan 4, 2005 10:51 pm 
Oliver Boehmer (oboehmer)Jan 5, 2005 3:15 am 
Chris CappuccioJan 5, 2005 10:15 am 
Actions with this message:
Paste this link in email or IM:
Paste this link in email or IM:
Atom feed for this thread
Paste this URL into your reader:
Subject:[c-nsp] MPLS, L2TPv3 Layer 2/3 VPN Network OptionsActions...
From:Oliver Boehmer (oboehmer) (oboe@cisco.com)
Date:Jan 5, 2005 3:15:15 am
List:net.nether.puck.cisco-nsp

One thing I've been curious about but haven't had the time/gear to lab is whether MPLS VPN (vpnv4 BGP/LDP) will play nice with the "confederation per pop" style of BGP confederations. We're currently a mix of partially meshed and multiple route reflectors, and it's getting to be a bit of a mess to maintain.

If you want to run MPLS-VPN in a BGP confederation setup, you will essentially build an Inter-AS MPLS-VPN network.

If you run a common IGP across all confed sub-AS', you don't need to worry about BGP next-hop, the only thing you need to disable on your confed-eBGP borders is the automatic RT filter, otherwise it will drop the vpnv4 updates (unless you configure all the VRFs on this router).

Not sure, but I'd rather clean-up the RR set-up than introduce confeds.. running MPLS in your core enables you to totally separate control and forwarding plane (i.e. remove BGP from your P devices) which makes RR placement very easy and straight-forward.

oli