05 5 14

Nate, Michael, Eric, Craig, Chris, Dominick, Satish, Jan, Kanika, Susan, Nick, Ryan,

Priorities :

1)  Updated tag so that next PID jobs have:
— proper preselection
— Improved memory usage
— Improved database monitoring
2)  Coordinate getting PID input files via CVMFS
3) PID for cosmic data  for data with 4 diblocks or less (Dominick)
(Make available for developer testing as they run)
4) PID for GENIE (~1000 files validation sample)    (Nate and Eric)  
(Run more if you can run them offsite)
(Make available for developer testing as they run)
5) Re-run reconstruction for cosmic data post 4 di block era data  (Dominick) 
(Then PID and CAF for this data)
6) PCHits for calibration stream for runs later than 14802   (Satish)  
7) “ulimit” some PID jobs with Jon’s database monitoring to understand crashes.  (Dominick)
8) CAF FTS issue (Michael)
9) CAF generation (Dominick) 
Question: Do we wait for LEM to make CAFs or do we make two versions?

PID update:
— Dominick posted slides:
— averaged calibration constants were not applied for much of the recent data reconstruction.
—> MC success rate: 25% (Friday evening)
—> Data success rate: 82% (Sunday afternoon)
— Michael will help with CAF FTS trouble shooting?
— cosmic data failure rate:

Restarting PID:
- base memory and nu_e preselection (CosRej already done a Reco step)
- Take out CAF re weight step here —

— GENIE:  Nate – jobs running.  (took couple of hours –> conflicts between Gavin’s test and Kanika’s updates)  
- Failure rate -> ??? 16+ hours.
- Writing to the wrong area, trying to move files
- We are currently not running the nu_e preselection (we should put it back in) — we should put it back in.
- Using an MC test release. Empty test release – minimize changes from Gavin. Should not do it in test release.
— Eric – some problems submitting jobs

Plan for PCHits (Satish): Jobs running…

New ART version:
— Who attended stakeholders meeting?

Ideas for database memory issue: Need better info on the crashes.