[j-nsp] L2VPN Active Route Selection
Stacy W. Smith
stacy at acm.org
Thu Nov 3 11:04:52 EDT 2011
Hi Zaid,
Apparently I sent this answer only to Eric instead of the rest of the list.
--Stacy
> -----Original Message-----
> From: Stacy W. Smith [mailto:stacy at acm.org]
> Sent: Wednesday, June 29, 2011 12:32 PM
> To: Eric Van Tol
> Subject: Re: [j-nsp] L2VPN Active Route Selection
>
> Eric,
>
> FYI, I think I figured out the answer to this. Seems the path
> selection algorithm for the L2VPN address family is completely
> different from the normal path selection algorithm (Section 9.1.2.2
> from RFC4271).
>
> The L2VPN route selection algorithm can be found in Section 3.4.1 of
> http://tools.ietf.org/html/draft-kompella-l2vpn-vpls-multihoming-02.
>
> --Stacy
On Nov 2, 2011, at 3:12 PM, Zaid Hammoudi wrote:
> Eric,
>
> Did you figure this out? I am having the same problem except with VPLS...
>
>
> Thanks,
>
> -Zaid
>
>
> On Mon, Jun 27, 2011 at 10:51, Eric Van Tol <eric at atlantech.net> wrote:
>
>>> -----Original Message-----
>>> From: Stacy W. Smith [mailto:stacy at acm.org]
>>> Sent: Monday, June 27, 2011 12:34 PM
>>> To: Eric Van Tol
>>> Cc: 'Juniper-Nsp'
>>> Subject: Re: [j-nsp] L2VPN Active Route Selection
>>>
>>> Do you have anything configured under [edit protocols bgp path-
>>> selection]? Specifically, do you happen to have "cisco-non-
>>> deterministic" configured?
>>>
>>> --Stacy
>>
>> No, I do not. The BGP config on the PE is pretty bare - no options
>> besides the family:
>>
>> admin at lab.router# show protocols bgp | display inheritance
>> local-address 10.18.20.46;
>> group 65000-dr {
>> type internal;
>> family inet {
>> unicast;
>> }
>> family inet-vpn {
>> unicast;
>> }
>> family l2vpn {
>> signaling;
>> }
>> export ibgp-car_to_dr;
>> peer-as 7784;
>> neighbor 10.18.20.72;
>> neighbor 10.18.20.73;
>> }
>> group 65000-vpn {
>> type internal;
>> family inet-vpn {
>> unicast;
>> }
>> family l2vpn {
>> signaling;
>> }
>> neighbor 10.18.20.10;
>> }
>>
>> I thought about configuring 'bgp-always-compare-med' or
>> 'cisco-non-deterministic', but figured it would be pointless, because the
>> MEDs are all equal, and thus, the options wouldn't have any effect.
>>
>> Thanks,
>> evt
>>
>> _______________________________________________
>> juniper-nsp mailing list juniper-nsp at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/juniper-nsp
>>
> _______________________________________________
> juniper-nsp mailing list juniper-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/juniper-nsp
More information about the juniper-nsp
mailing list