Re: [mpls] Comments on draft-ietf-mpls-tp-cc-cv-rdi-01

[David Allan I <[e-mail]>]

hi Lothar:

Thanks for the suggested clarifications. We'll update the document accordingly.

As for tunnel, we quite deliberately tended to steer away from any multiple LSP constructs as we considered that out of scope of the document. That is more of a survivability framework or draft-wiengarten topic. We are focused on the granularity of an LSP or bi-directional LSP pair (associated or co-routed). At the moment I do not see a reason to introduce tunnels into the draft...

hope that helps
Dave



________________________________
From: Reith, Lothar [mailto:[e-mail]]
Sent: Friday, July 23, 2010 2:22 PM
To: David Allan I; Greg Mirsky; John E Drake; [e-mail]; Annamaria Fulignoli; Sami Boutros; [e-mail]; Siva Sivabalan(msiva); David Ward; [e-mail]; [e-mail]
Subject: AW: [mpls] Comments on draft-ietf-mpls-tp-cc-cv-rdi-01


Dear Authors and All,

please find some comments below:

Issue 1:   inconsistent use of the abbreviation "CV"

 in the title the subject starts with "Proactive Connection Verification..."

However, in the abstract the meaning of the abbreviation CV is introduced as:

- Proactive Connectivity Verification (CV)

But in the section 2.1. Terminology the term CV is introduced as follows:

CV: Connection Verification

Suggested resolution: The abbreviation CV should be used in an unambiguous way throughout the document.



Issue 2: Clarification of the difference between the terms "Connectivity" and "Connection"

As the apparently ambigous use of the abbreviation "CV" shows, the terms "connectivity" and "connection" may be used synonomously in some cases.  However, I believe the concepts "Connectivity" and "Connection" are different.

Suggested Resolution: Provide a reference to where these terms are defined, and double check if the usage of these terms is always aligned with these definitions.

This should also help the reader in understanding the difference between the terms "connectivity", "connection", and "path", which are all used in the document, and perhaps also to the term "tunnel" in the meaning "MPLS TP tunnel" as related to issue 3 below:



Issue 3:  I am missing any mention of "MPLS-TP tunnels", even though an important construct within MPLS_TP is a connection which is provided across a working and a protection LSP.

Suggested Resolution: Consider adding a section "applicability to MPLS TP tunnels",  or add some wording in some other way that explains, if cc and cv are affected by the state of an MPLS TP tunnel being traversed, e..g when the tunnel switches from the working LSP to the protection LSP.

Best Regards, Lothar