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

Re: Network Configuration BOF - comments to IAB




As to whether a WG should be chartered, I think that's up to the operators. If they're not going to deploy it, we shouldn't build it, and what I would hope to hear is a strong message from the operator community that this specifically is what they want. Given that, I think the proposal is a reasonable phase I of the project.
The questions to the floor addressed to the operator community were relevant, and
there was evident operator interest.

Obviously operators have a fair amount of invested capital in CLIs both in mindspace
and in script space, and I'm not altogether sure that the XML approach will give a truly
standard (and vendor independent) form of config management in terms of standard
syntax and semantics, or whether the approach will at least provide a 'lock and load"
form of config management that improves upon the ratty scripts we have not that
attempt to impose common config frags updates across thousands of routers that
we use today.

To me the effort has the potential to make the operational support engineer's life
better, and potentially no worse, and that's a Good Thing.

I'm not at all convinced that the MIB to Schema thing is all that useful - we have to live
with the SNMP dogfood for many years yet as this is deeply embedded in most,
if not all operational environments (pity really as many vendor's MIBs for new
features are slow, poorly implemented and randomly documented, but thats life too,
and I understand the reasons why this is necessarily the case, and part of it is
because as operators we don't make MIB quality a valued part of our purchasing
decisions, so we get what we pay for. Of course the danger is that if we also view
xml-based conf in a similarly poor vein and don't place a tangible value upon its
quality of implementation, maybe it will suffer from similar benign neglect).

But I'm optimistic (I've got to be) and in that vein this is for me a productive effort
in the operational management that may yield useful outcomes.

Geoff