The MPLS-OPS Archive

[Truman Boyes <truman@suspicious.org>]

Hi Nathan,

MP-BGP can attach two new extended communities to routes that are  
redistributed from OSPF.

1. OSPF domain identifier
2. OSPF route route type

So with these attributes and the MED in BGP, you can preserve OSPF  
routing information across MPLS/BGP vpns.

Have you looked at using sham links to tie together your OSPF  
network? Specifically, using a sham link between VRFS on the 2 PE  
routers. OSPF will then create an adjacency between the two VRFs, and  
your ospf network will have metrics that consider the route across  
the two PE's.

Changing admin distance behavior is scary and should be avoided at  
almost all costs. I have seen terrible things happen in networks  
where operators had different ideas on preferred protocols (ie. LDP  
and RSVP routes).

Cheers,
Truman


On 12/03/2006, at 1:44 PM, Nathan wrote:

> Hi,
>
> I've got an annoying problem. I have a vrf that uses OSPF
> internally. To begin with, 2 PE routers conversing using
> classical MP-BGP. Then as I added links, I got a problem:
> packets coming in to a PE preferred to go out towards a CE
> linked to that same PE, even when it was much more efficient
> to use the other PE router to access a different CE. I traced
> that to the fact that IBGP links get a higher administrative
> distance, 200 instead of 110 for OSPF, so even if the OSPF cost
> on the connected CE link is 10000 and the OSPF cost on the
> CE link connected to the other PE is 10, the administrative
> distance wins out (the OSPF cost between the PEs is as far a
> I can see not taken into account at all, not that I care much
> about it).
>
> I looked at modifying the administrative distance, but that's
> system-wide, which would break loads of things.
>
> What I did was to set up a VLAN on other interfaces of the two
> PEs, then ran OSPF over that, no problem.
>
> Now I have to add links on a PE that does not have a convenient
> second interface for VLANs, just a single tag-switching
> interface and a lot of CE links.
>
> "OSPF Support for Forwarding Adjacencies over MPLS Traffic
> Engineered Tunnels" looks promising, but the document doesn't
> mention "vrf" once, and apparently it only exists on S/SXE/SBC
> trains (I'm running 12.2T and I don't even know what the S
> trains correspond to, do you need a different license? I know I
> use lots of features of 12.2T that aren't in 12.2 mainstream).
>
> I just want to pass the OSPF metric unaltered from one PE to the
> other . . . (unaltered or with the PE-PE cost added in, I don't
> really care).
>
> What should I do?
>
> Thanks for any help!
>
> Nathan
>
> -------
> The MPLS-OPS Mailing List
> Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
> Archive: http://www.mplsrc.com/mpls-ops_archive.shtml

-------
The MPLS-OPS Mailing List
Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
Archive: http://www.mplsrc.com/mpls-ops_archive.shtml