Thread bottlenecks such as contentions and deadlocks can slow your application or bring it to a halt. The Profiling and Logging perspective includes views and tools that can help you detect and resolve these thread problems.
You can use the following views to detect thread bottlenecks:
When working to detect thread bottlenecks, you often select profiling resources in the Profiling Monitor view.
You can also investigate thread bottlenecks using custom probes that you design in Probekit.
You can use the Thread Analysis profiling set to collect data about thread bottlenecks in your application.
In the Thread View, you can see a thread execution history for all threads in your application, plus explicit indications of deadlocks and contentions. In the Call Stack View, you can see stack frames for all the threads of the selected profiling resources. In the UML2 Object Interactions view, you can see the sequence of calls during the profiling run.
For more information, read Detecting thread bottlenecks.
In these views, you can see different formats of the raw data on which the Thread View is based.