[OPS-NM] RE: [IETFMIBS] comments on draft-natale-snmp-mibs-to-ontology-00

["Natale, Bob" <[e-mail]>]

 
Hi Dan,

You are correct in your perception that the proposed work *might* not
be appropriate for an IETF WG, esp. according to our traditional
objectives.  It is a proposal, in part, to move us more directly into a
broader interpretation of the relevance of core IETF O&M work (as
represented by MIBs in particular).  On the other hand, it might also
be seen as an extension of the MIB to XML work, NETCONF, XCAP, and
possibly others.  I will try to make that case during the BOF.

To answer your specific points:

1. I will provide more concrete explanations during the presentation
(and plan to post an updated Draft prior to the session...will e-mail
the link when it is ready).  But note that I do not claim to be an
expert in the set of SOA/WS mgmt related technologies that need to be
considered...I have a growing degree of familiarity with what seem to
be the key ones, but explicitly need the collective community expertise
of interested parties to make firm commitments to an optimal subset.

2.  Yes, "ontology" was used simply as a placeholder for the outputs of
the conversion process.  Since then I have been able to learn a bit
more, and get helpful feedback from some key SOA/WS mgmt developers, so
I can definitely narrow things down better for the BOF.  I strongly
doubt that we will be targeting ontologies as the outputs of the
conversion process at this time.  To give a preview, it is more likely
that resource models expressed as SML-IF documents better reflects the
kind(s) of things we would want the conversion process to produce.
(But see note above about need for community expertise to reach such
conclusions.)

3.  Yes, there was no intent to suggest that the WG (if approved) would
produce any tools.  Rather, the primary objective is to define the
specific rules for converting SNMP MIBs to SOA/WS mgmt artifacts.  I
firmly believe that such a set of rules is essential to both the user
and the developer mgmt communities and that the IETF is the right body
to specify such rules.  So, it is the conversion methodology itself --
not the tools that might use the methodology -- that I am asking the
IETF O&M area to standardize.  (Indeed, there are some open source
tools/projects that we might be able to leverage both to guide this
work efficiently and to test the methodology effectively via existing
tools along the way ... Some examples are the Eclipse COSMOS and Apache
Muse projects ... About which I will provide more details in the
updated Draft and at the BOF.)

Also, Juergen raised some specific questions in a (much) earlier
posting, and I will address those directly in the BOF presentation as
well.

Cheers,
BobN

-----Original Message-----
From: Romascanu, Dan (Dan) [mailto:[e-mail]] 
Sent: Wednesday, March 14, 2007 5:29 PM
To: [e-mail]; [e-mail]; IETF MIBs; MIB Doctors
(E-mail); [e-mail]; Netconf (E-mail); [e-mail]
Subject: [IETFMIBS] comments on draft-natale-snmp-mibs-to-ontology-00


I have a few comments related to the proposals in
draft-natale-snmp-mibs-to-ontology-00. While I appreciate the problem
space and recognize its importance, it is far from clear to me how this
work could be structured in a future Working Group in the OPS Area. In
order to help the discussions in Prague I would make the following
observations: 

1. The document relates to SOA-based/Web services management. These
concepts are quite broad, and it would be useful for the seek of the
discussions in Prague if Bob could provide a reference (or more but not
many) that points to what he had in mind when using this terms in the
document

2. The term ontology that is being used here to describe the results of
the conversion process is broad, and defined by a few examples in the
text. We should probably discuss rather sooner which of the output
formats that would be dealt with by a future IETF activity

3. Defining data conversion tools and validation tools is something
that
has not been done yet in the IETF, to the best of my knowledge. They
are
however defined as 'primary outputs' for this effort. As I am not sure
that I understand how the future IETF documents would look like, I
would
suggest that we make some more concrete proposals, maybe examples that
other have issued in this space. 

Dan

 


_______________________________________________
IETFMIBS mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/ietfmibs

_______________________________________________
OPS-NM mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/ops-nm