[rbridge] IS-IS pseudonodes, nicknames, and rooted trees
Russ White
riw at cisco.com
Tue Jan 9 06:42:15 PST 2007
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
> For RBridges, I think the answer to my question is that it does not
> matter -- I *think* the same
> tree would be calculated whether you chose as root the pseudonode or the
> DR. So it's probably better
> to say that pseudonodes do not request nicknames, and cannot be
> specified as the root of trees,
> and would never appear as the ingress or egress RBridge. I don't think
> it matters whether the
> endnode information appears to be advertised by the DR or the
> pseudonode. I think we can let the IS-IS
> people decide that detail.
I would agree with this. The main issue we might have is how to handle
the pnode advertisement by the DIS. I think this is okay, the
connections are all still there:
A---A'---B---C
If A' is the pnode, and A is the DIS, then A will be advertising:
o An LSP with it's own information, including its nickname and ID
o An LSP with the information form A', which would not include
a nickname
When C calculates its SPT, I assume that it will build the the link from
B--A', and A'--A, but without A' having a nickname, will this work? I
think it will, just wanted to think it through to make certain. The next
hop towards some destination, for B, could be A, but the local tree
shows this as through A', which should resolve in the forwarding table
down to the next hop A, just as it works today.
:-)
Russ
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFFo6nHER27sUhU9OQRAk4xAJ9eIma4KEKc2lCuSqoV7hJotwsfDACg7WX1
6Gp2uiB/usRhMjZsdcIQxOE=
=roMg
-----END PGP SIGNATURE-----
More information about the rbridge
mailing list