CNMP CPU Utilization Bug in OpManager
I read in your forums that the CPU Utilization, when reported via SNMP and HOSTRESOURCE-MIB, is supposed to be the average utilization of all CPUs/cores in the system.
I'm running OpManager v 6013 on a Windows Server 2003 SP1 x86 host. When monitoring CPU Utilization on a remote Windows Server 2003 SP1 x86 host with two (2) Intel quad-core processors I can confirm that what is being reported via OpManager is NOT the average. Or, at least it's not an ACCURATE average.
As a test, I pegged one of the cores with a single thread running in a tight loop on the quad-core box being monitored. All seven (7) other cores were essentially idle. Windows Task Manager reported an average of appx.. 13% during the test, whereas OpManager reported 95%.
This would appear to be a bug within OpManager. Please fix.
I'm running OpManager v 6013 on a Windows Server 2003 SP1 x86 host. When monitoring CPU Utilization on a remote Windows Server 2003 SP1 x86 host with two (2) Intel quad-core processors I can confirm that what is being reported via OpManager is NOT the average. Or, at least it's not an ACCURATE average.
As a test, I pegged one of the cores with a single thread running in a tight loop on the quad-core box being monitored. All seven (7) other cores were essentially idle. Windows Task Manager reported an average of appx.. 13% during the test, whereas OpManager reported 95%.
This would appear to be a bug within OpManager. Please fix.