Is my sample done?

There are a few places you can check to figure out if your sample is done processing.

Quicker (based on jobs):

  • Are the jobs still running? (Check Fifemon or run prodjob-summary on a gpvm machine.
  • Check the samweb project. If the number of files seen is equal to the number of files in the snapshot (and there aren't many error processes) then your sample is probably complete.

Better (based on sam definitions):

  • If you have made output definitions (How to make definitions), check that the number of files in those definitions closely matches the number of files in your input definition.
  • If you have made draining definitions (Construct a draining dataset), you can check that the number of files in the draining dataset is small.

What is "done"?

Generally, we do not require 100% of files to be complete to call a sample done, particularly for MC. Generally if we reach 99% complete then the sample can be called done, and in certain cases the bar is even lower. The only sample that must be 100% complete is the Far Detector Data.

What if it's not done?

Production Troubleshooting