Re: [PWE3] PWE3 new WG items?

[Nick DelRegno <nick.delregno at mci.com>]

A means of signaling/negotiating the QOS associated with a PW would be 
beneficial.  And doable.  We have seen implementations that already support 
two levels of QOS: one aggregate QOS for the PSN tunnel (usually a LSP) and 
discrete QOS for constituent PW Virtual Circuits.  I like this model.

Any transit LSRs in the LSP path with provide QOS on the aggregate LSP 
only, ensuring that the PSN tunnel has the proscribed bandwidth.  The PW 
QOS negotiation would be used at the PE function (traditionally at each end 
of the PW).  However, as we find new ways to apply these tools, the ability 
to signal the VC QOS and bandwidth becomes very interesting.  At some 
point, when we have multi-hop PWs (through stitching or other mechanisms) 
signaling the PW QOS instead of manually provisioning it would be helpful.

Whether we use CR-LDP, RSVP-TE or PW-specific extensions of 
"draft-ietf-pwe3-control-protocol-xx.txt" is immaterial at this point.  I 
agree that this should be a PWE3 WG item, as well as PW stitching and 
hierarchy.

Nick


_______________________________________________
pwe3 mailing list
pwe3@ietf.org
https://www1.ietf.org/mailman/listinfo/pwe3