02/01/2013 Minutes


Nick, Jan, Susan, Minerba, Kanika, Jarek, Himansu, Evan, Alex, Raphael, Philip, Ji, Ryan, Adam, Andrew, Jon, Chris, Nate, Mat, Peter, Robert

  • Action Item: Make the meeting a doc-db entry in future (DONE)
  • Announcement: Jan Zirnstein (UMN) is now taking over duties for tagging releases, maintaining the build and helping with software issues.

Raw to Root

Postponed to next meeting

File Requests

Include LEM processing on NDOS data. This will be done in future. Fast approaching Dallas stopped this happening already.
Process 2 orders of magnitude ND files for increased stat. studies. This can start now since it is to be reprocessed with S12-11-16 and is not reliant on any fixes/changes. Currently we have 131 ND FHC files, each ~500MB → ~65GB (very small sample).
The increase would put this to 6.5TB. The issue stopping us processing up to now was disk space. This has been addressed now. Andrew reported an expansion of /nova/app/ from 3TB to 8TB (technical limit for this array) and /nova/ana/ from 60TB to 95TB. A new disk is being added, /nova/prod/. Require a maintenance day for /nova/prod/. We also have/are recovering space.

  • Action Item: Phil will process increased ND stats in tag S12-11-16. This will then be reco’d in S12-12-12. Andrew will change permissions from novaana to novapro for all MC

Talk of increasing stats on nue_only files but forgone until an official request is made
by someone that is using them. Currently folks using these files are aware of a POT correction that is required. Robert H has circulated details of this and it is ~70x correction.
Will process all detector combinations through PID and CAF after a number of additions/fixes are made to the CAFMaker module (and some other fixes, more later). This processing will start on the scale of 1 week from now - the right people are aware of what they need to do.
Cosmics processing also held back until new CAF release (~1 week).

  • Action Item: Dominick will converge GD’s list with his own of variables and fixes for CAFMaker. He will present this at the Reconstruction meeting on Monday. GD to send hist list to Dominick first of course!

Spill Info

Currently only MC spill info being fed into CAFs. They look for the “generator” label, whereas Data spill info is under the “numispillinfo” label. This is symptomatic of development focus on MC up to now. Need to move towards MC/Data commonality. Information also carried by different objects MC → Data.

  • Action Item: Evan will work with Dominick to fix this issue, verifying with Tricia. Possibly need a switch for the next release to keep consistency across files. Next full production will need to fix this once and for all.
    Issue to be added to wiki to keep track of this.

MRCC issues

Robert and GD found genie and GEANT4 dependencies in EventDisplay coming from MRCC. Need to restructure MRCC to keep its data products separate from the source code in order to remove the reach of the dependencies. Kanika would also like to see MRCC in the production chain, though it needs to have files reprocessed as far back as ‘reco’, so this is not for the next CAF release.

  • Action Item: Kanika will work with Robert and GD to restructure MRCC. Kanika will also work on a strategy for implementing MRCC in the production chain.

Geom Service Issues

As brought up via email in advance and at nue, we need to fix up how our services decide the detector that is being used. Essentially we want to be able to use one .fcl calling services and not detector specific services.

  • Action Item: Chris has agreed to make the necessary changes with appropriate documentation after Jan tags this weekend (Superbowl release!)

/nova/data/ clean up

No objections to clean up.

  • Action Item: We will recover 51TB on /nova/data/ by deleting old release MC (don’t panic, it’s on tape!)

New people

New folks are starting their ramp up now. Please be patient.

Upcoming Meetings

SAM Metadata pow-wow on Tuesday 5th 9:30am -12 noon.
Scriptapalooza on Thursday 7th 1pm -5pm - half day "workshop" on designing and implementing effective scripts for physics data handling, production and analysis. You’ll have received email if you need to attend.
Andrew will attend the CVMFS seminar which conflicts on Thursday

  • Action Item: Flow chart and description of everybody’s production scripts needed before scriptapalooza. Production scripts to be placed in a common place. Gavin to bug necessary people on Monday.


Next production meeting is Friday Februay 15th 2pm