Hi,

On Thu, Mar 11, 2010 at 10:51:15AM +0000, Mateusz Blaszczyk wrote:
> you could probably do something like this
> 
> R1 ==(trunk)== Q1 -- R2 --(MPLS cloud)-- R3 -- Q2 ==(trunk)== R4

This might work as a workaround, or might not, depending on whether
R2/R3 would be willing to forward QinQ-encapsulated STP or not...

OTOH, we can life with "no working STP on R2->EoMPLS" for now, and 
making the whole topology even more complex is not going to help my
colleagues understand and maintain and troubleshoot the setup (something
to keep in mind).

> A solution would be to extended mpls towards R1 and R4 but that
> probably is not possible.

This is not possible here, because R1 and R4 might have SVIs in the
corresponding VLANs, and 6500 with LAN cards cannot do SVI+EoMPLS bridge
in the same box.

(Well, actually you can, by plugging a loop between two ports, one port 
being a switchport/trunk and the other port being the EoMPLS link.  But that
won't gain me much compared to what I have now - if the IOS combination
is broken with STP+EoMPLS, it will also be broken if that IOS is running
on R1...)

If I urgently needed the functionality today, I'd downgrade R2 from SXI2
to SXH3a - but I want $them to fix it, makes much more sense for everybody.

gert
-- 
USENET is *not* the non-clickable part of WWW!
                                                           //www.muc.de/~gert/
Gert Doering - Munich, Germany                             [email protected]
fax: +49-89-35655025                        [email protected]

Attachment: pgpG4NlSU7DGu.pgp
Description: PGP signature

_______________________________________________
cisco-nsp mailing list  [email protected]
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Reply via email to