The MPLS WG Archive

["Thomas D. Nadeau" <tnadeau@cisco.com>]

> -----Original Message-----
> From: Natale, Robert C (Bob) [mailto:bnatale@lucent.com] 
> Sent: Thursday, June 19, 2003 10:30 AM
> To: jcucchiara@mindspring.com
> Cc: 'MPLS WG'; tnadeau@cisco.com; 'Ash, Gerald R (Jerry), 
> ALABS'; Wijnen, Bert (Bert); 'Loa Andersson'; 'Lai, Wai S 
> (Waisum), ALABS'; 'Chung, Li-Jin W, ALABS'; 
> ppvpn@nortelnetworks.com; 'Van der Linde, Harmen, ALABS'
> Subject: RE: WG last call on LSR and LDP MIB modules - setting dates
> 
> 
> Hi,
> 
> Apologies for keeping the long cc: list....
> 
> > -----Original Message-----
> > From: jcucchiara@mindspring.com [mailto:jcucchiara@mindspring.com]
> > Sent: Thursday, June 19, 2003 8:37 AM
> > 
> > ...
> > My read of the draft is that the authors are indeed requesting
> > changes to the LDP protocol.  One example is the
> > discussion on sessions being enabled/disabled.  As Tom points out, 
> > this is not in the MIB because it is not in the protocol.
> 
> Yes, and we all understand (I hope) the futility of that.
> 
> However, these exchanges do raise several very good points:  E.g.,
> 
>    - Isn't it a good idea to have a fuller understanding of
>      how a technology should be managed *before and/or during*
>      protocol development?

	Yes. 
 
>    - If so, perhaps:
> 
>       - Each protocol development WG should have a Mgmt & Ops advisor?
> 
>         (The "MIB Doctors" service is valuable but seems to tend to
>         focus on SMI compliance (and similar issues) to the exclusion
>         of MIB sufficiency relative to the management requirements of
>         using the technology at hand.)

	Yes, I agree. 
 
>       - The Mgmt & Ops Area should publish a set of critical generic
>         management requirements for all protocol development WGs to
>         refer to as they do their core work.

	BUT, those requirements should be used hand-in-hand
with those produced by the WG. Also, the WG may have a good
reason to not use some of those requirements and should be
free to do so if there is consensus to do so.
 
	--Tom


>         (Yes, everyone knows how important management and operations
>         are :-], but everyone also knows how much that is just given
>         "lip service" in the segments of the real world that don't
>         include real users.)
> 
> > ...
> > A question was made privately to one of the authors at the
> > Atlanta IETF that perhaps a MIB which stores LDP counter information
> > in a history table (such as is done in SONET) could be 
> useful to them?
> 
> This is a good example of the kind of generic feature and 
> "re-use" that I am talking about above.
> 
> Cheers,
> 
> BobN
>