I have some 511 purepaths lets say /PRService.
Response time ranges from 100 to 700 ms. Purepath comparison shows they are structural the same, but have timing differences.
17 of them are marked with a single long sql query problem. If the purepaths are structurally the same is there a timing component in the long sql query problem detection.
I am analysing these in the context of stuck threads and was hoping to identify problematic methods. Seem to me either all 511 have the same pattern or non of them have it if they are structurally the same?
Can you explain how this problem detection works and wether or not I can rely on the problem detection as accurate when the purepaths are structurally the same?
The problem detection pattern looks at the purepath and matches based on a set number of patterns that we have seen (n+1, long running query, too many threads) and takes a best guess at highlighting that for you.
For further troubleshooting your issue, my best guess would be that the bind variables or the data lookups are different. I would look at capturing bind variables or further instrumenting the method to highlight the types of data that it is looking up.