Showing results for 
Show  only  | Search instead for 
Did you mean: 

This product reached the end of support date on March 31, 2021.

Difference between sum and count aggregation


Hi Team,

Can anyone please explain me the difference between sum and count aggregation with examples. I actually need to know number of hogging threads present in an application for a time interval of 30 min which we will specify in incident rule. Can you please tell me what aggregation should I use here.




DynaMight Leader
DynaMight Leader

Have a look at the answer here:

The explanation from andi would also apply for your case.

Dynatrace Certified Master, AppMon Certified Master - Dynatrace Partner -

Thanks Patrick. I have already seen this post but not able to get clear picture so posted a question once again.

Which measure typre are you using exactly for the "hogging threads"?

Dynatrace Certified Master, AppMon Certified Master - Dynatrace Partner -


Please find the attached screenshot. The measure we are using for hogging thread is that only. There is one aggregation in this measure and also we have to specify one more aggregation in the incident rule which we create by adding this measure. What is the difference between this two and also what exactly should we specify.

I would still guess sum and count is the same here. Could you try just charting both aggregations ob the measures in a chart and check if they are the same?

Dynatrace Certified Master, AppMon Certified Master - Dynatrace Partner -

Inside the measure under aggregation there is no count thing. But need to know actual difference between sum and count. @Patrick H. @Babar Q. @Patrick H.

Dynatrace Guru
Dynatrace Guru

Sum is the sum of the numerical value that the measure is tracking, whereas count is the count of PurePaths the were involved when calculating that measure. So if your measure is for web requests and 2 PurePaths involved 4 web requests each, then the sum would be 8 and the count would be 2.

In your case I would recommend using the sum aggregation.



What kind of aggregation do we need to use for JVM thread count?

Hello @Bhaskar G.

Use the average aggregation for the thread count.




Thanks @Babar Q.

We are seeing an issue with WebSphere JVM thread count, where the dashboard we created gives us the accurate value but the alert throws us a different number. dashboadwhere-the-threadcount-is-less-than-35.png

Any input on this is very helpful. Thanks in advance.


The measure we are using is this

We created an incident rule and added this measure in that rule..