I were try to monitoring a process which is called by batch file.
And I've captured the purepath.
But the process would have a long execution time and more than 10,000 PurePath nodes so that I need to adjust setting for the PurePath.
I try to Setting PurePath timeout first.
After I commit the setting, the PurePath would captured more then 10 mins and I can see the PurePath when the process is running.
However, When the process done, the PurePath disappear immediately.
I've try to adjust filter to find the PurePath but still failed.
I also try to reproduce the issue in my test environment.
And I adjust the same setting in my test environment to make sure if it caused by the setting
The result is : The setting is Okay.
Because the issue happened, I didn't adjust the upper limitation of the nodes.
I have no idea and worried about the issue.
Any advise will be appreciated, thanks.
Both of those settings should be tweaked with max purepath nodes as well. But be aware that this may produce some overhead for Appmon server and should be used carefully.
About missing purepaths. It is possible that it is there but there is missing execution time because of time outing. In such case it may be visible as transaction with Response time 0ms. If it’s completely missing, you should check agent / collector and server logs. Maybe there will be any clue. I was successfully recording purepaths that took 45 minutes some times ago using Appmon so in general it should work. If you have to many nodes in purepath, consider if instrumentation isn’t too deep. Sometimes you may see some unwanted methods on PP which may be issue as well for such long transactions.
Thanks for your help!
May you tell me more about PurePath completeness timeout and Maximum PurePath age ?
I've read the page but still confused.
As my understanding:
The "Maximum PurePath age" is about the total running time of PurePath. It would stop recording the PurePath when the Maximum PurePath age reached.
And the "completeness timeout" only occurred when one of the method has long response or even no response.
Am I miss-understanding about these two timeout?
This means that even if you will not reach timeout while for example pp is waiting for response from asynchronous task, but you will reach maximum pp age, server will stop collecting this pp.
Thanks for your kindly help!
I checked the log you mentioned, and I found the message below:
Skipping a PurePath: Live PurePath: agentId=ab51e408, tagId=0 from realtime analysis and session storage because it is already too old. This PurePath will not be analyzed, as we cannot adapt measures that far in the past.
But I don't understand what does it mean?
Do I miss some settings? (I added the completeness timeout to 3600s and the max nodes to 20000.)