PCM+ 2.1 unreachable switches when using sampler traffic monitor

[J Vesterdahl]

Since I've changed my traffic monitor from stats to sampler on my 2824 switches, I notice that now and again, one or more switches become unreachable for a few minutes. When it happens, I can log in to the switches by telnet or web, so it looks like they're ok after all.
Is this due to the sampler traffic monitor? I did not have any problems using stats. My max load on the network is 3%. Switch FW is 08.71. PCM+ is v2.1 and not upgraded.
Any ideas?

[Matt Hobbs · Accepted solution]

I haven't seen that before, but first things first I would recommend you update to the I.08.98 - http://www.hp.com/rnd/software/J4903898.htm

This release now includes some 'show sflow' commands for troubleshooting sflow.

What I think will help is, in PCM under Preferences, try increasing the Status Polling Retries.

My best guess is that sflow is putting an additional load on the switches and they are not always responding in time for PCM.

Don't forget to assign points to any replies you receive.

[J Vesterdahl]

Okay, increasing the retries works so far.
Thanks.
It puzzles me that a switch with only 3% traffic load is disturbed by the traffic monitor, though. What happens if the traffic load is above 50% ?
Anyway, I am going to upgrade the firmware network-wide, but not three days before my vacation ;-)