The MPLS WG Archive

[Curtis Villamizar <curtis@workhorse.fictitious.org>]

In message <4B6D09F3B826D411A67300D0B706EFDE0115CCD9@nt-exch-yow.pmc_nt.nt.pmc-
sierra.bc.ca>, Shahram Davari writes:
>  
> Curtis,
>  
> I think you are misreading the draft. It doesn't say that there
> shouldn't be any OAM tools for ECMP, PHP, Mp2p. All it says is that
> special considerations and complexities are involved in those cases,
> and it outlines the issues that needs to be considered while designing
> an OAM tool.
>  
> So I don't see your point as a valid point to discard this draft, and
> I am sure many other WG members agree with me.
>  
> Yours,
> -Shahram


I think the draft as is could be submitted as an informational draft
submitted by Dave outside the work group.

As a framework for OAM it is a diversion from progress rather than a
step toward progress.  It would be enough to give a simple statement
that any form of OAM which simply injects traffic at ingress and
counts traffic at egress is insufficient in many circumstances and
perhaps briefly enumerate the circumstances rather than describe them
in detail.  The framework can then outline requirements agreed to in
the WG and if it follows other framework documents it may touch on
solution approaches concentrating on those that work within the entire
framework (work in all cases, meet the largest set of requirements).
Mention can be made of those approaches that work in a limited context
but the overall emphasis should not be on evaluation of the
implications of network topology on a specific solution that has been
rejected by the WG.

Curtis