Re: Adoption of draft-vkst-bfd-mpls-mib

[Jeffrey Haas <[e-mail]>]

Adrian,

On Sat, Jun 02, 2012 at 05:31:11PM +0100, Adrian Farrel wrote:
> The document claims to extend BFD-STD-MIB but there is no citation for that
> module and I can find no active document on the charter page that includes that
> module. Has something timed out? Is the cart in front of the horse?

Yes, it's timed out.  It's in the enviable position of being also ready for
last call, but only once the authors republish it.  (Which they've been
prodded to do a few times in the last couple of months.)

> And a meta-question (arising from recent pushback I have been getting for MIB
> modules within the IESG)...
> Notwithstanding MIB modules being a deliverable of the WG, you should only work
> on them if there is a clear desire to build them into product and use them in
> deployments. You should not spend cycles making MIBs just because they are in
> the charter, and certainly not just because "here is a chance to publish an RFC"

As anyone who follows MIBs is aware, I'm hardly the last person to say
"let's do a MIB, just because". :-)

> Lastly, a number of objects in the module are read-create. This implies that the
> plan is for the module to be used for (or at least capable of being used for)
> configuring BFD behavior, not just monitoring it. I would like the WG to be
> certain that this is function it wants. But that question does not need to gate
> adoption if the WG chooses to work on the I-D.

As part of the IETF MPLS-TP work, I believe we (the IETF) are expected to
provide interfaces to configure that feature.  While I'll admit to not
playing enough in that field to know what the latest conversations have been
on this topic, the current choices are either SNMP or netconf.  I believe
there is more support for SNMP at this point.

It's probably worth prodding the MPLS chairs to offer a public opinion on
the matter.  I've cc'd them on this and will also forward the original email
to them.

-- Jeff