[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Target times for MIB Doctor Review

Wijnen, Bert (Bert) wrote:

I am here in an IESG retreat about improving IESG/IETF performance.

One of the things being discussed is that MIB doctor review (one of
the situations where some ADs put the doc in "expert review" status
as the substate in the I-D tracker) can take enourmously long.

The main reason is that it is often difficult to find a reviewer, and
then the doc sort of by defaults ends up in my queue (which is too
long already).

So the suggestion is that we would like to have a target of MIB
Doctor review to be done within 30 days of the request.

I want to hear from you how to deal with that?
If we cannot find a good solution, then the only alternative that
I can see is that we will get MIB documents onto IETF Last Call
and onto IESG without the normal MIB Doctor review, and so the
OPS-NM AD (for now that is me) will only check for fatal errors.

The result might be that we end up with MIB documents in much worse
shape than what we have seen over the last few years.

What do we think about this?


I don't have time for MIB reviews myself, so technically
I shouldn't even be on this list, but...

FWIW, I struggled with this problem a great deal at Cisco.
MIB reviews take too long from everybody's perspective,
and a volunteer review team that gets no real credit for the
job it does is a recipe for disaster.  Only a total newbie
will benefit from a simple GUI that hides the SMI.
The problem is not the SMI.  The problem is the 847
CLRs (I mean guidelines) that come with the SMI.

The answer is tools development.
Short of that, we now have a comprehensive MIB review
guidelines RFC, and multiple MIB compilers that produce
detailed error messages.

You can select a severity level that all MIBs must pass,
which is less than "super squeaky clean".  Point people
at the docs.  Use MIB Doctors for advice and verification
that a particular MIB module passed the designated severity level.

This means you have to decide not to make review comments
like "I don't like the name of this object" or "Redesign this part
of the MIB module this way".



Bert





Andy