LogEndToEndEvent.js/ps1 CPU impact
the interal health check script runs on all monitored computers at the exact same time. Please find a way to reduce the VM host CPU overhead this causes.
Confirmed by MS Premier Support. This rule does not do anything else than to log the event. The monitors using the event are disabled by default. It is totally safe to disable, and as long as you have *any* other event collected by SCOM you have confirmation of event collection work
Is it possible to get an official comment on this? I could disable the following rule http://systemcenter.wiki/?GetElement=Microsoft.SystemCenter.AgentManagement.LogEndToEndEvent&Type=Rule&ManagementPack=Microsoft.SystemCenter.2007&Version=7.3.13142.0
But i'm not sure that's good, but how often is event log collection on an agent broken?
Hi Kevin, yes, we noticed the cpu impact on our hosts. Managed to track it down to that particular script using perf.mon etc. over a few hours. Not a problem for a single VM but for a host you will see CPU wait time etc caused by this as all VM's want to run at the exact same time.
Added a few screen shots. one for a VM and one for the host who struggle the most
Kevin Holman commented
Are you actually seeing CPU impact here in practice, or in theory? Because the simple writing of an event is VERY low CPU impact. Like difficult to measure, low.
Could probably change:
<DataSource ID="Scheduler" TypeID="System!System.SimpleScheduler">
To(will need to allow some overrides or hardcode the values instead):
<DataSource ID="Scheduler" TypeID="System!System.Scheduler">