Re: ATM connection management

[Joan Bessette <[e-mail]>]

Re: ATM connection management <Bob Stewart>
>Rather than prioritize reasons, I'd rather restate a general principle.  Every
>MIB object must pass the scrutiny of "so what."  You must be able to answer
>why the network manager cares, and what he's supposed to do about it.  You
>don't have to list every use of a given object, but you should be able to give
>at least one clear, strong example of its value.  "Interesting" isn't
>sufficient. 

I believe we have to hold these MIBs up to a higher standard then the "so what"/interesting
test, if these MIBs are going to be accepted as a standard then they must conform to the 
framework.  From an NMS perspective I'm interested in how standard MIBs relate to each 
other.  I would like to see some standard guidelines/model in place that would force the MIB 
structure to follow a defined schema.  I want my MIB model to reflect the communications/
applications model, so that I can pick and choose the MIBs I implement based on the actual
technology.  

Re: ATM connection management <William Manning>
>Two thoughts and some free association.
>
>	1. the orad wg is beginning to have actual network operations folks
>           review drafts for operational impact.  instead of guessing what 
>	   a network manager might think, you could ask orad to review.
>	2. there seems to be a feeling on those folks that are doing draft
>           reviews, that the SNMP MIB descriptions are the last things selected,
>	   since they are SOOOOO cryptic. or maybe just dry.
>
>        F. Interesting might be sufficent, if only to force something to be 
>	   so unwieldy that it doesn't get used. :)  Just what are the hidden
>	   adgendas?
>
It seems that a lot of the MIB objects are related to IEEE/ISO standards, which
is good, but these usually convern the extreme case rather then the norm.

Regards,
JB

Joan Bessette 
Motorola-Codex 
617-821-7854
[e-mail]