Re: Challenges for the BGP MIBv2

[Jeffrey Haas <jhaas@nexthop.com>]

On Wed, Sep 29, 2004 at 10:01:03PM +0200, Juergen Schoenwaelder wrote:
> We prefer to let IANA manage the OID name space and not to let WGs to
> do their own little OID name space management. There were cases where
> this approach did not work too well. And it does technically not really 
> matter where a module is registered. See also the MIB review guidelines
> draft-ietf-ops-mib-review-guidelines-03 section 4.5.

Given recent answers in this thread, I can understand why the mount
point of the extention is irrelevant.  However, from a conceptual
standpoint and also given the issues raised in the recent Early IANA
Allocation Internet-Draft, wouldn't an IANA managed namespace in
the BGP MIB make more sense, especially since these are extensions?
After all, the extension modules can't be used independantly of the
BGP MIB and placing them under the BGP tree would reflect this dependency.

I don't have a strong opinion on this.  However, the primary object
in the MIB review guidelines is that the WG managed the space
rather than IANA.

> This seems to answers my question. You will probably have to rewrite
> most of the current MIB anyway to make it address family independent.

This is one of my remaining questions that had been addressed to
our previous MIB doctor.  When re-writing the MIB, since most objects
will require outright replacing, will this be placed under the
existing mib-2.15 as new objects or should I expect a new mib-2
mount point?

Knowing the expected answer to this allows for a small amount of
re-use of a few existing objects.

> Juergen Schoenwaelder		    International University Bremen

-- 
Jeff Haas 
NextHop Technologies