mantid-profiler

Use psrecord and Plotly.js to profile a Mantid workflow.

It monitors CPU and RAM usage, and reports the time span of each algorithm (currently, only algorithms are recorded).

Usage

To profile the SNSPowderReduction.py workflow:

python SNSPowderReduction.py & python path/to/mantid-profiler/mantid-profiler.py $!

The script attaches to the last spawned process, so you can also use the profiler if you are working with MantidPlot:

./MantidPlot & python path/to/mantid-profiler/mantid-profiler.py $!

psutil
You need to build Mantid with the -DPROFILE_ALGORITHM_LINUX=ON CMake flag to get the timing output from the algorithms.

After running on the SNSPowderReduction.py workflow, the profiler produces a profile.html file to be viewed with an internet browser.

You can interact with a demo profile here.

Controls:

-h, --help show this help message and exit
--outfile OUTFILE name of output html file (default: profile.html)
--infile INFILE name of input file containing algorithm timings (default: algotimeregister.out)
--logfile LOGFILE name of output file containing process monitor data (default: mantidprofile.txt)
--diskfile DISKFILE name of output file containing process disk usage data (default: mantiddisk.txt)
--interval INTERVAL how long to wait between each sample (in seconds). By default the process is sampled as often as possible. (default: None)
--noclean remove files upon successful completion (default: False)
--height HEIGHT height for html plot (default: 800)
--bytes Report disk speed in GBps rather than Gbps (default: False)
--mintime MINTIME minimum duration for an algorithm to appear inthe profiling graph (in seconds). (default: 0.1)

viztracer creates similar information for generic python software

Neil Vaytet, European Spallation Source