Hello Manuel,
I've been reading PCE works for about a year. I cannot find any document
that explains how to avoid loops in interdomain PCE path computation. I
don't know if there is a specific doc that covers this issue.
Do you know something more?
I assume you mean how to avoid loops created by re-entering a domain (loops
within a domain are simply the job of the computation algorithm).
So far, little attention has been given to the selection of the sequence of
domains to be crossed by an inter-domain path. Although it is recognised as
a potential element to computation techniques it is also known that initial
deployments will be severely gated by administrative and commercial
limitations (e.g., inter-AS peering agreements) such that the sequence of
domains for a path will be known in advance.
As draft-ietf-pce-brpc-05.txt says:
The PCE-based BRPC procedure applies to the computation of an optimal
constrained inter-domain TE LSP. The sequence of domains to be
traversed can either be determined a priori or during the path
computation procedure. The BRPC procedure guarantees to compute the
optimal path across a specific sequence of traversed domains (which
constitutes an additional constraint). In the case of an arbitrary
set of meshed domains, the BRPC procedure can be used to compute the
optimal path across each domain set in order to get the optimal
constrained path between the source and the destination of the TE
LSP. The BRPC procedure can also be used across a subset of all
domain sequences, and the best path among these sequences is then
selected.
Note that the brpc technique itself would initially appear to be vulnerable
to looping between PCEs, but the use of a pre-determined list of domains
means that this cannot happen.
And, of course, if you apply the per-domain approach, the sequence of
domains is also known in advance.
More interesting, perhaps, than deliberate re-entry into a domain, is the
selection of the sequence of domains so as to avoid re-entry. This function,
however, sails very close to the wind with regard to our current scope. The
charter says:
The PCE WG will work on application of this model within a single
domain or within a small group of domains...
and clearly the selection of the sequence of domains is less interesting if
the number of domains is small.
But, nevertheless, brpc could be used to perform this function (as described
in the quoted text above) by examining the alternative paths that are
available from pre-determined sequences of domains.
Further, the original PCE architects did consider a "hierarchical
cooperative PCE" model whereby a parent PCE is aware of the topology and
connectedness of domains (but not of the content of the domains) and sends
computation requests to the PCEs responsible for each domain. Such a model
is quite close to that expressed in the ITU-T's G.7715.2 (even though the
domains have a somewhat different meaning in that case).
Regards,
Adrian
_______________________________________________
Pce mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/pce
