Chassis MIB. Multiple agents

[[e-mail] (Niels Ole Brunsgaard)]

[e-mail] writes:

>I don't understand what 5.6 below means. Does it mean that, in a
>particular chassis, I may have multiple agents that export the same
>instances of chassis MIB objects?  How/why would one implement this
>arrangement? Is it viable/useful for a chassis as big as Texas?

>If I am a management station, is this replication of the chassis
>information at multiple agents helpful? Confusing? How would I sort it
>out?

These are my reasons for wanting multiple agents in one chassis supporting 
the chassis MIB;

	1) The Chassis MIB provides pointers to the other agents in the
	chassis. There are no pointers the other way. Thus a MS cannot
	find out which agent provides the chassis MIB view, when managing
	an entity without a chassis MIB.

	2) Many of todays modular systems allow hot swapping of cards. Thus
	the chassis MIB view may disappear at any time. Instead of having 
	the remaining entites detect the disappearance of the chassis MIB 
	agent and then provide a new view, I find it easiser/better to have 
	multiple views from the beginning.

>Is it viable/useful for a chassis as big as Texas? 

I don't know.

>If I am a management station, is this replication of the chassis
>information at multiple agents helpful? Confusing? How would I sort it
>out?

I think its helpful for the reasons above. Since all agents provide 
identical values, I don't think its confusing and no sorting should be
needed. It's like eating at McDonalds. The location doesn't matter. They
all have the same menu.


	Niels Ole Brunsgaard.
	Cray Networks.
	Copenhagen, Denmark.