exchange 2000 problem false reporting in SIM

[Frank Fish]

We are getting pages almost daily that our exchange cluster has gone down. The exchange cluster never fails over and I get the following message:

engw2kpex1,User Threshold - Rising Alarm (rev. 3),Threshold Variable: cpqHoFileSysPercentSpaceUsed.1,Threshold Type: absSuppressFallingTrap,Current Value: 92,Threshold Top: 90,Threshold Owner: Insight Web Agent,Severity Level: warning,Description: D:EX_Application [NTFS] disk space

This seems to indicate to me that the D: drive is running low on disk space. However it has 9gb of 12 gb free. Anyone have any ideas why this might be happening?

[David Claypool]

Is 'engw2kpex1' the name of the cluster or one of its members? If you view the disk utilization page for that member, what is the threshold?

[Frank Fish]

Thanks for the reply. PEX1 is the name of node 1 in our exchange cluster. The threshold is currently set at 95%. I have checked the D: drive to see if we are approaching this limit and it is never even close. Even immediately after we get the pages/messages. The strange part too is in addition to the message I posted it also states the cluster failed when in fact it has not. Thanks again for your help.

[David Claypool]

If as you say it is set to 95, right now it thinks it is set to 90 and the current value is 92.

My best suggestion would be to go to the device virtually or physically and go to

Control Panel --> HP Management Agents

Under the 'SNMP Settings' tab is a [Clear All Thresholds] button. Do that for each node and then set the thresholds the way you want them again and let's see what happens.

[Frank Fish]

Sorry, I think I misspoke a bit earlier. The thresholds are 90 (major) and 95 (critical). I will reset the thresholds as you indicated and see what happens.

[Frank Fish]

Claypool,

I reset the threshholds as you mentioned before. I then copied enough data to the drive so that it pushed it over the 90% level. We got paged with the threshold warning as we should have but we did not get the cluster failed messages. This works as it should it seems. When we get the false threshold reports they are always accompanied with cluster failed reports too.