Can one of the admins verify this patch?

The replay contains three runs to test all the fixes:

• Run 359762: Fix replay with GsfTracking no states are passing selection criteria
• Run 361971: Fix paused job with GsfTracking fix unprotected area
• Run 361468: Fix abort job with protection for njettiness for JetMET

@germanfgv @jhonatanamado Will the replay take too long or could it be fine?

@ebrondol this is way too big. 361971 and 361468 are 0.5PB of data. We can have a replay for 359762, which is a good pp run with all components in, as a general test.

We can also have reduced replays with the bigger runs, in which we only process the affected stream.

@ebrondol this is way too big. 361971 and 361468 are 0.5PB of data. We can have a replay for 359762, which is a good pp run with all components in, as a general test.

We can also have reduced replays with the bigger runs, in which we only process the affected stream.

Looks like a good proposal to me!

Is it possible to include the specific events that failed for these patches? I imagine not, so I guess we should just mention here any private tests that checked them.

test syntax please

@germanfgv CMSSW_12_4_11_patch1 is available on cvmfs:
/cvmfs/cms.cern.ch/el8_amd64_gcc10/cms/cmssw-patch/CMSSW_12_4_11_patch1

Can we start the replay(s) you mentioned above?

Yes, here you can follow the general replay with 359762
https://monit-grafana.cern.ch/d/t_jr45h7k/cms-tier0-replayid-monitoring?orgId=11&var-Bin=5m&var-ReplayID=221118101417&var-JobType=All&var-WorkflowType=All

hi @germanfgv , if I see the grafana plots correctly, the replay with 359762 is going well.
Are we waiting for it to completely finished before testing the runs with problems or we could start them already now?

@ebrondol about the other replays:
I started the 361971 replay at the same time. It already finished successfully. Here you can see the monitoring:
https://monit-grafana.cern.ch/d/t_jr45h7k/cms-tier0-replayid-monitoring?orgId=11&var-Bin=5m&var-ReplayID=221118104713&var-JobType=All&var-WorkflowType=All

I also tried to start the replay with 361468, but discovered that we no longer have the streamers. Checking the deletion logs, it looks like we added the run to the exclusion list a bit too late. This means we cannot test this with replay, but we'll have relay on the tests done with the job's Pset.

After the successful test, this was merged into production here #4784