Monitoring uses a large amount of CPU on client
Hi,
I am currently evaluation the product and have an issue with the amount of CPU usage required on the target domain controllers.
Ever since we installed OpManager we have been getting frequent alarms due to the processor queue length thresholds being broken; currently set as the default of 7. Further investigation has led me to conclude that these alarms are occurring when OpManager is polling the domain controllers. OpManager seems to utilising 100% of the CPU for about 5 minutes every 15-20 minutes.
Some additional information -
Why is so much CPU usage required and why is the pattern every 15-20 minutes when the monitoring interval is set to every 5 minutes?
Regards,
Doug
I am currently evaluation the product and have an issue with the amount of CPU usage required on the target domain controllers.
Ever since we installed OpManager we have been getting frequent alarms due to the processor queue length thresholds being broken; currently set as the default of 7. Further investigation has led me to conclude that these alarms are occurring when OpManager is polling the domain controllers. OpManager seems to utilising 100% of the CPU for about 5 minutes every 15-20 minutes.
Some additional information -
-
- The usage of 100% is only on our older servers. Newer servers still show the same pattern but the CPU usage is less.
- None of the monitoring intervals are set to 20 minutes. The DC monitoring interval is 5 minutes.
- I have check Microsoft's site and ensured that WMI logging is set to error only and not verbose
- If I use perfmon on the DC and delete the DC from OpManager we do not get this pattern. i.e the problem is due to OpManager
- I initially thought that the problem was the size of the event logs ( and in particular the security log). I have changed the auditing so that very little is written to the event logs
Why is so much CPU usage required and why is the pattern every 15-20 minutes when the monitoring interval is set to every 5 minutes?
Regards,
Doug