Last week I learned something from a lab utube, speaking about disk alerts, the statistics interval of 15 minutes so he set his alert to trigger at 16. So I tried to use this new found knowledge to look at CPU alerts. Originally the alert was to evaluate every 5 minutes and alert when exists for 10 minutes. So what happens if I set the exists for 20 minutes?
I get an alert, development node. But the list of processes, here's the top 4:
WmiPrvSE.exe 18032 5.19 %
SWJobEngineWorker2x64.exe 22668 1.63 %
svchost.exe 896 0.45 %
System 4 0 %
This alert cleared after 10 minutes. This is outside of my exists time period but would only match up to the check for condition. Selecting the gauge it gives me the chart. Raw data should help me look at this? It's showing data only every 30 minutes which matches the statistics collection.
I'm having a real hard time with several aspects but really trying to get an understanding of the layers to make an effective monitoring system with valid/essential/etc alerts. Am I outside of the park or maybe even in the parking lot of the park? Searching to understand the relationships for this purpose has been difficult so I thought I'd ask and see if there was any good information out there that I'm just not finding/getting.
Thanks much.
