P14 Reprocessing: Meeting of 10-Dec-2003: 10:15-11:00 VRVS Desert.

Agenda

1. New Projects.
2. Corrupted Files.
3. Status of Sites.

Minutes

Participants

Topics

1. New Projects (Mike):
   Try to assign remaining projects according to rate delivered so far. NPACI and SAR projects are reassigned.
2. Corrupted Files (Mike):
   What to do with corrupted input DSTs? Compare CRC locally computed with CRC stored in SAM. If it differs try to transport again, else notify SAM people to mark it as corrupted.
   Corrupted output files: First retry. If it still fails, don't store them into SAM!
   Thomas observed DSTs not located in SAM. In this case contact Mike with the filename. He will check whether the file can be recovered.
3. Site Reports:
   • Farm: Still doing reprocessing (expected for another week)
   • GridKa: Slowed down last Wednesday because of diskspace. Off since Saturday. Half a day of full runnning away from completion dataset. SAM station to be fixed after change of IP address.
   • Lyon: -
   • Nikhef: Glueing pieces together for EDG usage. Running on local PBS only. DSTs are stored in SARA (not announced to SAM).
   • UK: RAL and Viking (Imperial) storing DSTs directly to SAM (enstore).
   • WestGrid: Around 3.5M events done. Computing center is instable, so low efficiency. DSTs stored locally (not announced to SAM).
4. Estimated end of the processing: Beginning of January. Around 80M events are still to do.

Action Items:

Next Meeting