At 10:53 PM 11/26/2004, Curtis Villamizar wrote:
At least on MPLS FRR implementation provides a cost assigned per SRLG
and something called a bias which is another number space. A non-zero
bias in the path means that the backup is considered inadequate (the
local-protect-available bit is not set). Setting costs on the SRLGs
allows the operator to specify which SRLG are most in need of
avoiding, where node-disjoint may be specified as an SRLG, therefore
giving among ranking SRLG to the probability that the node goes down.
Being able to support the same in IP FRR might be nice.
The mechanisms would need to be different. There are 2 parts to the
problem - the first part is controllable (to some extent) like TE FRR and
the second part isn't, b/c it is just the SPT.
One can track the SRLGs and neighbors visited and make decisions based upon
that. Those decisions could be made based on weighting of the SRLGs.
It's not clear to me that it is necessary to go to this level of
complication. It also makes it difficult for any distributed decisions
where X needs to be able to determine what Y picked as an alternate or
whether Y picked anything.
> I think that if the point where a repaired packet re-joins the SPT is
> always a downstream path, then we could have the same
> property. Unfortunately that does limit coverage a bit.
You seem to be missing the point. If there is an unknown SRLG the hop
that you thought was downstream isn't really downstream anymore. That
is the basis for loop formation with unknown SRLG.
Curtis, think about this as a 2 stage problem. The first stage is do the
alternates cause a loop. The second stage is does the convergence to the
new topology cause a loop. We are working on mechanisms to handle the
second; for instance the path-locking with safe neighbors doesn't care
about whether the set of failures were expected.
For the first stage of do the alternates cause a loop, that can be handled
by requiring that the alternate rejoin the SPT at a point downstream of
where it left. This is very similar to TE FRR, where the backup rejoins
the primary always downstream of the failure.
Rtgwg mailing list