|
|
|
| ______________________________________________________________________ |
COMPREHENSIVE CPU PROFILING
|
- OptimizeIt's CPU profiler provides comprehensive, accurate data on CPU/time usage.
- OptimizeIt's CPU profiler offers unique features such as the option to use either an instrumenting CPU profiler or a sampling CPU profiler, to obtain the degree of details that is most efficient for profiling any kinds of Java programs, while maintaining a low overhead.
- The CPU profiler provides the option to measure either pure CPU (time spent during execution) or time (time spent during execution + inactive, waiting phases).
Instrumentation and sampling options for accurate tracking of performance bottlenecks (JDK 1.2 required)
- OptimizeIt is the only tool to provide both sampling and instrumentation profiling options. This allows you to tune CPU/time measurement to the most efficient level of detail for each test session while maintaining a low overhead.
- Instrumentation is recommended when profiling small amount of code. Sampling is indispensable to rapidly identify performance bottlenecks in large Java programs.
CPU/time usage graph
- The CPU/time usage graph allows you to rapidly identify time-consuming methods and corresponding lines in your source code.
- The CPU profiler displays CPU/time usage per thread or thread group. A graph shows for each thread or thread group the time spent in each method during the test session. The graph displays the tree of all method calls, time spent per method is displayed in percentage and absolute time.
- For each method in the tree you can automatically display the corresponding line(s) of source code with the source code viewer.
Automatic highlight of busy threads
- The CPU profiler automatically highlights threads that were busy during the profiling session, so you can rapidly spot threads that were most active. You can then select all threads or thread groups and see where time is spent within each thread with the CPU/time usage graph.
Hotspot display
- The Hotspot display lists methods where most of the time was spent. The time shown is the time the program spent in a method, no matter where the method was called from. The information is provided in milliseconds, percentage of time and invocation count.
- This feature is particularly useful to understand if a single method acts as a bottleneck and can be optimized to speed up all the tested features.
Source code highlight
- For each string representation, the responsible line of code is highlighted in the source code viewer, allowing you to track in your source code the lines responsible for performance problems.
- You can track performance problems in Java programs even if you don't have access to its source. Source code highlight is the only feature that is not available when the source code is missing.
|
|
