In one of the monitored applications, processes were found high at 70-75% and later normalized . Now am asked to tell what was the cause of that in the application but i do not know how to drill down to that problem! Kindly help
As you can see from the screenshot posted, the CPU is generated by a monitored process. Find out which one is it (from Host Health dashboard) then drill down into it (Process Health dashboard). See if anything points out -> increased transactions? garbage collection?
If not, try opening the Response Time Hotspots dashlet, add a timeframe filter for the time period when your CPU spiked, change the grouping to agent instead of agent tier -> can you see the agent responsible for the spiking process? -> click on the bar to drilldown to Method Hotspots (it will be auto-sorted by CPU Sum), identify the top methods responsible -> can you associate these with a certain transaction?
It's a bit of a vague question, but I think this is a good start.
Chris, no need for CPU Sampling.
From main menu on your left, expand Performance Analysis, then Response Time Hotspots, change timeframe to 19.11.2017 16:00 - 20.11.2017 20:00 change grouping (button in top right corner from Agent Group to Agent) -> is there a bar on the chart on the left labelled "_Concierge_Portal_243Admin[MS3_ugpttcoapp13]" ? If yes, click on that bar. It will drill down and open Method Hotspots for you. Here, look at the top methods listed, see if you can assign them to a specific transaction (use the tree view at the bottom of the screen to navigate through methods/services which call this method).
Hope this is a bit more clear.