 |
The hot spots view is conceptually similar to the CPU call tree view.
The information on the method nodes refers to the recorded payload names. For example, if an SQL statement was executed
42 times at a particular call stack with a total time of 9000 ms, this adds an event count of 42 and a time of
9000 ms to all ancestor call tree nodes. The leaf nodes show the payload named that are recorded by the probe.
The cumulation of all recorded payloads forms the call tree that shows you which call paths consume most of the
probe-specific time.
Even if sampling is enabled, JProfiler records the exact call traces for probe payloads by default. If you want to avoid this overhead, you can switch it off in the profiling settings. For more information on the payload concept, please see the corresponding help topic |
| | Most features of the CPU call tree view are supported in the probe call tree as well, including setting the call tree root, creating call tree analyses as well as method splitting and exceptional method runs. A chosen call tree root remains specific to the selected probe and does not propagate to other probes or to the CPU views. |