To start Performance Analyzer , type the following command in a terminal window:
% analyzer [control-options] [experiment | experiment-list]
You can specify an experiment name or a list. The experiment-list command argument is a space-separated list of experiment names, experiment group names, or both. If you do not provide an experiment list, Performance Analyzer starts and opens the Welcome page.
You can specify multiple experiments or experiment groups on the command line. If you specify an experiment that has descendant experiments inside it, all descendant experiments are automatically loaded and the data is aggregated. To load individual descendant experiments, you must specify each experiment explicitly or create an experiment group. You can also put an en_desc directive in an .er.rc file (see en_desc { on | off | =regexp}).
In reading experiments with descendants, any sub-experiments that contain little or no performance data are ignored by Performance Analyzer and er_print.
To create an experiment group, you can use the –g argument to the collect utility. To manually create an experiment group, create a plain text file whose first line is as follows:
#analyzer experiment group
Then add the names of the experiments on subsequent lines. The file extension must be erg.
When Performance Analyzer displays multiple experiments, data from all the experiments is aggregated by default. The data is combined and viewed as if the data is from one experiment. However, you can also choose to compare the experiments instead of aggregating the data if you specify the –c option. See Comparing Experiments.
You can preview an experiment or experiment group by clicking its name in the Open Experiment dialog box.
You can also start Performance Analyzer from the command line to record an experiment as follows:
% analyzer [Java-options] [control-options] target [target-arguments]
Performance Analyzer starts up with the Profile Application window showing the named target and its arguments and settings for profiling an application and collecting an experiment. See Profiling Applications From Performance Analyzer for details.
You can also open a “live” experiment – an experiment that is still being collected. When you open a live experiment, you see only the data that had already been collected when you opened the experiment. The experiment is not automatically updated as new data comes in. To update, you can open the experiment again.