I have a few customers that have DCRUM-App Mon integration and so far based on my experience both products usually report similar/very close response time for operations (in DCRUM) and corresponding PurePaths (in App Mon). For example, a 56.7 sec slow operation in DCRUM should drill down to its corresponding PurePath with a 56 sec +/- 0.5 sec response time (or duration time if there are a lot of asynchronous calls).
However last week a customer noticed that there is a significant difference for one of its operations:
From the above screenshots, the operation in question has a server time of 47.8 sec. However when we drill down to App Mon:
App Mon shows that this PurePath has a response time and duration time of 2.79 sec.
What caused the discrepancy?
Based on verbal communication with customer they didn't see a lot of async calls and the Response time = Duration, which is also 2788.07. So I am not sure what caused the discrepancy...
Chris, can you tell a little more about your setup? Which DC RUM version is that?
I think we would need a vdata/hpdata for such operation or an ADS report with related metrics. Plus diagnostics from AMD - have you checked if there are no packed lost? If there are, two hits might merge into one operation and render such confusing results.
Hi, are there no firewall / load balancer between place where AMD is located and first node monitored by AppMon (Web Server?). Often if you have AMD places before / after load balancer you can see discrepancy like this that helps in investigating faulty domain more precise than AppMon that obviously focuses on server-side monitoring.