Cell Relay Archive

[gsundar@home.com]

> > The other reason for SPVCs over PVCs is in link failure situations.
> > PVCs by definition do not reroute.  If the failed portion of the circuit
> > is over non self-healing facilities (not SONET or SDH), you're SOL. PVCs
> > have no re-routing mechanism - they depend totally on the underlying
> > SONET layer to provide fault tolerance.  Various manufacturers overcame
> > that "flaw" with proprietary means, and the ITU and ATM Forum
> > consolidated them into a standard (or recommendation, in ATMF-speak).
> > SPVCs, since they are SVCs, will attempt to re-setup the connection
> > first over the failed link, and then via alternate routes.
> >
>
> soft pvc recovery from failure is the responsibility of the "owner" of
> the soft pvc and is optional and requires that the owner attempt to
> reinitiate setup of the connection.  the recovery scheme for soft pvc's
> has whatever means are made available in pnni (which means crankback
> procedures, to my knowledge).

Again : this recovery is a connection setup time recovery, whereas
the self-healing SONET/SDH is not confined to just connection setup
time. As someone pointed out, we need proprietary methods to simulate
SONET kinda recovery on Soft Pvcs (by having a proxy user who will
check the sanity of the link even after it's setup - as per a previous
posting about one such mechanism on GDC switches ?).
So aren't we talking about different things : SONET self-healing is
a 'real time' (in the sense of being monitored continuously *after*
the Setup) whereas in the case of Spvc or Svc's it's only during
Setup time ?

- Sundar

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own