SamWiseMinutes2 6 14¶
Nate, Adam, Gavin, Craig, Eric, Jan, Chris, Andrew, …
(Fermilab all-hands meeting may still be in progress)
------------------------------------------------------------------------
— Adam: How close are to being completely reprocessed Raw2root
— Last night about 9000 had not yet been processed.
— about 5000 jobs waiting…
— Over 90% done.
— Near keep-up mode
— Who can start doing keep up.
— Need someone to take over from Adam.
— Production account,
— Gavin has been using the Raw2Root files in SAM to fill in files for neutrino search.
— Neutrino data being used well into January. Played a part in that.
— Eric ran the Near Detector beam and cosmic successful onsite and offsite.
— reconstruction
— Still need to figure out how to do overlay.
—> grab multiple secondaries to mix with a primary.
—> suggestion to file a ticket with the Artists.
— Adam met with Chris and Andrew to discuss LEM mixing
— So far, script edits a fcl file…
— ArtSamWrap wraps the “nova” command
— Hard part: don’t know what file you are going to get until run time.
— May need to communicate with the artists on this too
— Jan — we have a new tag.
—> PID tests are now ready to go. Fcl updated.
—> algorithm uses to environmental variable to find the PID libraries.
—> Now CVMFS location variable is also
—> Eric volunteers to test it offsite today.
—> Includes metadata module update — also NOVA.standard flag.
— ArtSam wrap is part of IFDHART, but we have a committed version as well.
—> Need to make sure that we don’t diverge.
—> Nate needs to forward his changes to Art.
— Nate: PCHIT files into SAM today for testing.
— New category in FTS file — hierarchy
— Single submit can’t be more than 6000 jobs – if you have multiple submits you might get around that ???
—> Need to understand this better.
—> Limited memory on head node.
— Eric: MC question:
— out of 100,000 events only about 2000 passed selection in cosmic. (one in 50 spills has a cosmic in it)
— send around the statistics
— More FD MC cosmics will be requested.
— Eric: database small tests using validity context tables (reco). Work perfectly.
— People doing studies with old release may have things broken when Jon updates the validity table.
— Dataset naming convention and dataset librarian