[c-nsp] L2VPN Pseudowire Redundancy
Ivan Gasparik
ivan at ig.sk
Mon Nov 3 04:01:42 EST 2008
If the goal is hardware redundancy (to ensure working VC when one of
the PE's fails), you can create one VC and configure it using two
separate loopback ip addresses. You will assign a loopback interface
to PE1 and PE2 with the same ip address. The other couple of PE's -
PE3 and PE4 - will share another ip address on their loopbacks.
Every PE will have VC configured using these ip addresses, you will
choose one of each couple as primary and configure IGP to pick ip
address of the primary PE's and propagate it across the backbone.
In case of primary PE failure will IGP do its job - propagate the
loopback ip address of the backup PE and allow LDP to establish new
session between working PE's.
Ivan
On Monday 03 November 2008, Rakesh Hegde wrote:
> How about creating two psudowires , PE1- PE3 and PE2-PE4 ? This
> will give you two logical point to point connections between SW1
> and SW2 and at the same time take care of device (PE) failure .
> STP,by default, will take care of the redundancy. You may also
> want to use UDLD and/or PAGP or LACP to provide end to end link
> status.
>
> -Rakesh.
>
> On Sun, Nov 2, 2008 at 2:07 PM, Arie Vayner (avayner)
<avayner at cisco.com>wrote:
> > I would suggest that you treat these 2 parallel PW's as 2
> > separate L2 connections.
> > Each connection would be handed over to the end customer
> > separately, and the customer can run STP end to end between their
> > CE's.
> > This way the failover between PW1 and PW2 would be based on
> > CE-to-CE STP
> >
> > Alternatively, if the customer is using L3 CE's, then its just 2
> > parallel L3 links...
> >
> > Arie
> >
> > -----Original Message-----
> > From: cisco-nsp-bounces at puck.nether.net [mailto:
> > cisco-nsp-bounces at puck.nether.net] On Behalf Of Mateusz B?aszczyk
> > Sent: Sunday, November 02, 2008 21:40 PM
> > To: giesen at snickers.org
> > Cc: cisco-nsp at puck.nether.net
> > Subject: Re: [c-nsp] L2VPN Pseudowire Redundancy
> >
> > you would have to land these xconnects on VPLS instance.
> > so add 4 more devices that would be your N-PEs with VPLS instance
> > and your current PEs would become U-PEs connected to the rest of
> > the MPLS cloud with 1 xconnect to the "active" N-PE and backup
> > xconnect to the "standby" N-PE.
> >
> > But I am not sure it is possible on 7206.
> >
> >
> >
> > --
> > -mat
> >
> > 2008/11/2 Gary T. Giesen <giesen at snickers.org>:
> > > I'm not sure if this is possible, but maybe someone can give me
> > > some input on how to best achieve this.
> > >
> > > I'm labbing EoMPLS using 4x 7206 VXR. I'd like to create a
> > > fully redundant pseudowire (from the provider persective).
> > >
> > > The idea is to put two PE routers at each end of the pseudowire
> > > (with a common VLAN at each end shared through a switch), so
> > > that I can fully lose a PE router and the VC still stays up.
> > >
> > > The topology looks like this:
> > >
> > > [PE1]
> > > [PE3] CE1 --- [SW1] ---< > [MPLS CLOUD] <
> > > >--- [SW2] --- CE2 [PE2] [PE4]
> > >
> > > I've tried a number of ways using xconnect-peers and backup
> > > peers (per
> > > http://www.cisco.com/en/US/docs/ios/12_0s/feature/guide/fspseud
> > >o.html ), and it works great when I only have redundancy on one
> > > end, but as soon as I add the 4th PE, nothing works anymore.
> > >
> > > When I add the 4th PE router, PE1 forms a VC with PE3, and PE2
> > > forms a VC with PE4, when in reality I should only ever have
> > > one VC formed at any given time, and PE2 should never form a VC
> > > with PE4 until PE1 or PE3 goes down.
> > >
> > > Does anyone have any suggested configurations?
> > >
> > > Regards,
> > >
> > > GG
> > > _______________________________________________
> > > cisco-nsp mailing list cisco-nsp at puck.nether.net
> > > https://puck.nether.net/mailman/listinfo/cisco-nsp
> > > archive at http://puck.nether.net/pipermail/cisco-nsp/
> >
> > _______________________________________________
> > cisco-nsp mailing list cisco-nsp at puck.nether.net
> > https://puck.nether.net/mailman/listinfo/cisco-nsp
> > archive at http://puck.nether.net/pipermail/cisco-nsp/
> > _______________________________________________
> > cisco-nsp mailing list cisco-nsp at puck.nether.net
> > https://puck.nether.net/mailman/listinfo/cisco-nsp
> > archive at http://puck.nether.net/pipermail/cisco-nsp/
>
> _______________________________________________
> cisco-nsp mailing list cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
More information about the cisco-nsp
mailing list