Skip to content

PerfMonComps: Add a minimal job to the CA configuration that demonstrates what PerfMonMT can do

Alaettin Serhan Mete requested to merge amete/athena:master-perfmonmt-ca-run into master

This MR makes the PerfMonCompsConfig.py runnable for demonstration purposes. It'll process 100 events in which we execute one algorithm (CpuCruncherAlg) that uses 100 +/- 1 ms/event of CPU time and another algorithm (LeakyAlg) that emulates "leaking" 40 KB/event. Upon successful termination, the user should end up w/ a report from the service that will include (among other things) results along the lines of:

[...]
INFO Number of events processed:        100
INFO CPU usage per event [ms]:          100
[...]
INFO Leak estimate per event Pss:       40.00 KB
[...]

Merge request reports