Folks,
I was wondering whether anyone has tackled the problem of identifying
a failed switch in a Cisco 3750 switch stack (or similar).
Until recently we had switches that had their own IP address & were
monitored as individual entities by Intermapper. Now Cisco 3750s are
being deployed. The stack has one IP address & within the stack
individual switches can be added/removed/fail etc but the single IP
address keeps on responding (thus giving the illusion that all is
well when it isn't !)
In dealing with this type of problem on other devices, we have been
able to find a "Global Status" variable somewhere in a manufacturers
MIB, which will tell us whether the device is fully functional,
degraded etc. It is far from clear (to me) whether such a variable
exists buried in the multitude of MIBS that a 3750 supports.
We have also considered
a) monitoring traps emitted by the switch.
b) with the latest switch firmware - monitoring the interswitch
links. We make these as small as possible on a map - but because they
are 'not hidden' - then intermapper will warn when one of these
interfaces fail. (ie we use the standard Cisco process & memory pool
probe & from the interfaces window hide all but the interfaces
entitled "StackSub-St1-1","StackSub-St1-2", "StackSub-St2-1" etc)
Any ideas, thoughts or comments welcome
Thanks & Regards
Stewart
--
Stewart Harris Email: [EMAIL PROTECTED]
The University of Sheffield Tel: (0114) 222 3074 (direct)
Corporate Information & (0114) 222 2000 x23074 (switchboard)
Computing Services Fax: (0114) 222 1188
285 Glossop Road
Sheffield
S10 2HB
England
____________________________________________________________________
List archives:
http://www.mail-archive.com/intermapper-talk%40list.dartware.com/
To unsubscribe: send email to: [EMAIL PROTECTED]
