Skip to content

Fix longest processing time calculation

Stewart Martin-Haugh requested to merge smh/HLTMPPU:fix_longest_processing_time into master

I noticed that the "longest processing time" printout at the end of the job looked odd:

Number of:
  events:     80
  - accepted: 80
  - rejected: 0
Average durations (ms) / fraction of total time:
  l1Result:   21.05 / 0.0040607
  process:    5158.15 / 0.995045
  - accept:   5158.15 / 0.995045
  - reject:   0 / 0
  send:       3.525 / 0.000679998
  total:      5183.84
Longest times (ms) :
 - L1 result  : 106
 - Processing : 2.51748e+08

With the fix, this is back to normal now:

Number of:
  events:     80
  - accepted: 80
  - rejected: 0
Average durations (ms) / fraction of total time:
  l1Result:   18.9875 / 0.00368167
  process:    5135.86 / 0.995841
  - accept:   5135.86 / 0.995841
  - reject:   0 / 0
  send:       1.425 / 0.000276307
  total:      5157.31
Longest times (ms) :
 - L1 result  : 33
 - Processing : 29450

Merge request reports