Cell Relay Archive

[nobody <nobody@web.onecall.net>]

Thierry Arnoux wrote:
> ["nobody"] wrote:
> > Now, the question is:  do you need to have separate instances
> > of these for each DS1, or can you do a meaningful aggregation?
> > If the former, then provide one row for each DS1 ifTable entry
> > (which is not compliant with RFC 2515, but should work);  if
> > the latter, provide one row for each ATM  interface.
> 
> An aggregation can surely be done, but the only meaningful solution
> is to have separate instances of the atmInterfaceTCEntry for each DS1.
> As you write, it is not compliant with RFC 2515, and this is my problem.

Understood.

> > P.S.  You should send questions/comments like this
> > to the IETF AToM MIB working group mailing list.  See
> > http://www.ietf.org/html.charters/atommib-charter.html
> > for subscription info.
> 
> This surely helped - I could find the following mail in the archives
> of the IETF AToM MIB working group, which described exactly my problem !

You probably had to do quite a bit of legwork to find that message.

> However, it was not followed up. I presume that I have to wait for
> the next revision of RFC 2515 to have it officially solved...
> 
> I will maybe post in the IETF AToM MIB working group mailing list
> anyway.

I think that you should do that.  The WG is now finishing up on three
important specifications ... last call closed today on the updates to the
SONET-MIB (RFC 2558bis) and the closely allied PerfHistTC-MIB (RFC2493 bis),
and last call expires tomorrow on the initial version of OPT-IF-MIB (optical
interface MIB).  One of the next new work items will be advancing (or
recycling) the ATM-MIB (RFC 2515).  Whether this work is undertaken will
depend in part on whether there is sufficient interest (and, possibly,
the availability of an editor).  So, please let your interest be known.
Things _won't_ move if everyone just waits for something to happen.

In the meantime, the suggestion made by Kevin Nolish, viz.

% The simplest way to correct the MIB would be to use the ifIndex of the
% underlying physical layer object, say Dsx1 interface, as the value for
% the index of the atmInterfaceTCEntry table.

is what I would implement if I were in your shoes.  It's the obvious
solution, and when RFC 2525bis appears it's almost certainly what will
be in it.  I would be surprised if it's not already a "de facto"
standard adopted by many IMA vendors.  Asking on the AToM MIB list
is a very good way to find out if that is really true.  Hope to see
you there!

nobody