Project

General

Profile

09/11/2014

We had an important discussion regarding direct access to blue ark disks from the grid. CD would like to limit direct read/write access to the blue disks. Direct access to /nova/app is critical to NOvA under the current protocol. Concerns were also expressed regarding what this would mean for reading and writing to /nova/ana, /nova/data, and /nova/prod. Matthew and Craig will discuss with Andrew and instigate a wider discussion once the proposed plan is understood.

Attendance

Matthew, Chris, Kanika, Satish, Craig, Nate, Michael, Jeny, Paola, Dominick, Gavin, Gareth.

News (Matthew)

  • ProductionScripts setup with permission for me and NOvA pro. Upon Nate’s suggestion it was decided to include three subdirectories one directory for fcls, one for submission scripts and one for grid scripts.
  • Calibration want tricell & activity processed through PC hits.

PNFS usage discussion (Craig)

Craig opened a ticket to clarify this, the response was it there is 200 TB of space for NOvA and it sounds like we can write there, the files don’t get archived to tape and after a period of inactivity oldest files will get removed. They recommend that we clean up after ourselves.

Action Item: (Craig?) will update the wiki or post a link.

Dataset definitions from make_sim_fcl update (Nate)

Nate’s setup consists of a single job for fcl creation, dataset creation & submission. The first pass of this was done on Monday. It was decided that it would be a good idea to test this before the Simulation tag. This requires a tag on CMVFS. We need a new tag to deal with the rise time issue anyway.

Raw2root keep-up audit (Jeny)

Jeny identified a large number of files that weren’t RAW to rooted. She will continue to look into the nature of these files to determine if we need to process them.

PChits keep-up (Paola, Satish)

Paola has Satish’s scripts and they’ve defined steps to complete the hand-over process. Paola is currently automating the scripts & making them novapro-able. ND calibration files will start being processed when hand over happens. It is likely that these files will be “urgent” in a few weeks time.

Reco keep-up (Jeny, Dominick)

Dominick started his back processing jobs, he found, fixed and back ported some fixes. The files have already starting to appear & some are in bluearc already. All ND data will be done today. Then he will work with Jeny on handover.

Action Item: How do we want to archive our keep-up? By run period? By month?

Sim validation (Gareth)

Simulation want their validation to run as standard as part of production. Several options on how to do this were discussed. An email discussion will start after the meeting to workout the best plan of action.

New ART tag and are we ready to tag for Simulation? (All)

Need to fully understand the implications of the rise time problem. Apparently this is baked in at cal hit time. Chris has committed a fix for the data, but it looks like this fix didn’t make it in time to be in the keep up processing.

Discussion of user tools for SAM analysis (Gavin et al.)

Postponed to be first on agenda on monday.

AOB

  • Dominick is away 17-27th