General info

  • The hps-java Release 3.11.1
  • JAR: /u/group/hps/hps_soft/hps-java/hps-distribution-3.11.1-bin.jar
  • An example JSub script XML file
  • All Steering files taken from JAR (i.e. they are used with the option -r)
  • Logs: /lustre/expphy/work/hallb/hps/data/physrun2016/pass1/logs
  • The Detector is HPS-PhysicsRun2016-v5-3-fieldmap_globalAlign
  • Blinded pass: Only 10% of files are reconstructed, files: 0, 10, 20, etc
  • Files on work disk /work/hallb/hps/data/physrun2016/pass1

 

Change in the storing files on tapes

In  This pass, and probably in all future passes, individual outputs (except recon.slcio) were not written into the tape, but instead for each run and category, all files are tarred, and instead a tar file is sent to tape.

The maximum tar file size is chosen to be 20Gb, if total size of all files is more then the max_size (20 Gb), then files are distributed among more than one tar files, in way that each individual tar file has not more the 20 Gb.

For example for "dst"s of run 7800 two tar files are created "tar_0_hps_007800_dst.tar" and "tar_1_hps_007800_dst.tar"

 

 

 

Additional information

Job submission scripts are organized in a way, to avoid hard coded values as much as I was able, and almost no hard coded values are in the code.

Tables below define the whole structure of the pass, and job scripts get all the necessary information from two files "PASSINFO.dat" and "PASSOUTPUTS.dat"

Content of PASSINFO.dat

 

Content of PASSOUTPUTS.dat

 

  • No labels