The MPLS WG Archive

Cell Relay Retreat>MPLS WG Archive>month:2000-Jul> msg00449



[Date Prev][Date Next][Thread Prev][Thread Next]  
  [Date Index][Thread Index][Author Index][Subject Index]

questions on draft-kompella-mpls-te-mib-00.txt

  • From: Kireeti Kompella <kireeti@juniper.net>
  • Date: Sat, 29 Jul 2000 01:26:55 -0700 (PDT)
  • Cc: mpls@UU.NET, te-wg@UU.NET

Hi Joan,

> Draft-kompella-mpls-te-mib-00.txt has objects which
> are not specific to TE per se, but to monitor general information
> about LSPs  -- (specifically objects under the mplsInfo
> branch and about half of the objects in the mplsLspList (Table)).

The LSPs in the TE MIB are TE LSPs.  They are named -- in fact,
they are indexed by name, although this is a point of contention.
They expected to have an ERO and other properties that are more
natural to TE LSPs than to LDP LSPs.  But your comment raises an
issue that is easily resolved: why are LDP and BGP mentioned as
signalling protocols for what are meant to be TE LSPs?  If this
is a problem, I'll take them out of the list.

> Is there any plan to separate this document into 2 documents?  There
> is a natural division for LSP objects in a separate document
> from TE objects, and I would like to see this happen.
> MPLS LDP implementations, for example, could support the LSP objects but
> could not support the TE objects. 

You probably know that there is an LDP MIB in the MPLS WG :-)
I see the two as complementary.

> Secondly, could you tell me what the plans are for defining objects
> which would tie LDP Session manipulation (as defined in the LDP MIB)
> and this this MIB?  The prime example would be Session teardown 
> which happens as a result of user interaction with the LDP MIB.
> How do you see the LDP MIB and this MIB (in different working groups)
> being co-ordinated?   Are you also going to define objects which allow
> the LDP Sessions and this MIBs LspLists to interact?  I would like
> to understand what course of action is going to be taken  
> with regard to LDP and the LDP MIB since this MIB clearly involves LDP.

The excellent issues you point out don't arise if the TE MIB only
talks about TE LSPs (set up by RSVP/CR-LDP).

> Lastly, is the TE working group going to be developing MIBs and/or PIBs? 
> In my opinion this is not clear from the working group's charter.

The TE WG falls under O&M; I would think that TE MIBs would be
right up their alley.  I'll suggest that the chairs make it
clearer in the charter.

As for PIBs: good question; you'll have to ask the chairs.

Kireeti.