System vs user and process CPU criteria?
Can anyone explain to me the criteria dynaTrace uses for ascertaining host 'system', 'user' and 'monitored process' CPU Usage metrics? We have one environment that consistently will have a 'system' component of the CPU Usage that is 2/3 or more of the total. Even though the overall usage is not alarming, dynaTrace produces a lot of built in alerts because (by default) it seems to think anything above 15% 'system' usage is suspicious. I searched the manuals and can't find anything regarding how it determines what is 'system' vs. 'user' CPU use.
And yes, by the way, I do know how to adjust this threshold to mitigate the alerting, but we're wondering why the 'system' component is so high in this particular environment.
Hi Pete
What OS are you running on?
Windows server 2008, or 2012.
I reached out to engineering to learn more about the windows API we use to get this time. I know we use system level APIs for that - but I want to make sure we tell you what method it is exactly so that we can point you to the MSDN Documentation.
