Monitoring processor thresholds through HP SIM

[Tekkie]

Hi all - Have you successfully used the option of monitoring processor utilisation on Proliant Windows server systems? I've managed to set my servers up to alert on a CPU threshold breach of 85% and HP SIM reports on this fine. But unlike the disk thresholds function there doesn't seem to be a way of resetting the alert. So as a result I still have 80 or so servers showing up in HP SIM gui as having processor issues even though they are currently nowhere near the threshold percentage. Any ideas on this?

[Rob Buxton]

Not sure if I'm seeing the same issue, from the SMH on one of our servers the Processor performance is shown as 100% on each of the two CPU's But if I look at the Processor Utilisation screen which doesn't have the sliding controls it reports around 6% usage which agrees with Perfmon.

So here, there seems to be an issue with the CPU performance reporting.

This is using 7.60 of the agents. The server also happens to be a DC.

[Tekkie]

Hi Rob,

It is a related issue because it is the foundation agent on the server that I believe is responsible for reporting on the metric. My thought was that now that theproceesor(s) currently are at normal levels maybe the agents should be stopped and restarted, hoping to reset the reported alerts in this way. However, on doing this it seems to clear the alert in HP SIM but if you return to HP SIM 5 minutes or so later there it is again!. Which kind of makes me think that the reporting is partially based on some sort of log rather than solely on realtime events.

[Tekkie]

not really a suitable solution but the only way to reset the alert is to move the threshold pointer up and down (or down and up!) and saving each change in between. The alert will disappear in the Insight Agent and eventually in HP SIM, until the threshold is broken again