Project

General

Profile

11/20/14

Attendance

Matthew, Chris, Dominick, Gareth, Nate, Kirk, Satish, Craig, Gavin, Ryan, Jeny.

News (Matthew)

  • MR Brem status - Nitin is travelling, so in the meantime Nate will top up the cosmic sample to the full 20k files and then later work with Nitin on the extra 2M events requested.
  • SAM for users status - Work has started on this. Jeny is the point person, Andrew is keen to have working meetings so that Robert and Marc can help out where necessary.
  • Subrun merging - Justin is in classes so progress on this will be slow (but that is ok for now as this is our third option for reducing the number of files).

SLF6 upgrade (Craig)

Craig called and talked to Andrew about the problem we've been seeing to do with the SLF6 upgrade (which so far has happened to novagpvm{01-03,09}). As a result of this Andrew communicated to CD to not update the machines until they have the ok from him. Andrew noted that the gpvm09 upgrade was done ages ago and there were no reported problems. Chris pointed out that this machine had been taken out of the normal rotation of nova-offline logins, so people simply weren't using it.

It was decided that Craig will send an email to the collaboration encouraging the use of these machines, but pointing out that there could be some small issues that will need to be communicated back to CD.

First analysis progress (all)

  • FD data - pid-part and lem-sum ongoing.
  • FD beam MC 14DB - RHC still being generated, not all datasets are quite full yet.
  • FD cosmic MC - top up will start on this soon.
  • ND data - Rapid turn around completed, and we'll wait on a new tag before processing anymore.
  • ND beam MC - all in Rapid turn around as above.
  • ND cosmics - will reco when we have time.

Ryan raised the question on how to fold in post shutdown data into what we’re processing. It was pointed out that to do this we'd need bad channels, a good runs list (not strictly required to add more data) and to extend our "random runs" MC conditions to cover the extra data. This question mostly affects the FD keep-up and the ND data. FD keep-up should start ~next week when a new tag is completed. It was decided to pick up this discussion at the all analysis meeting.

RHC files (Nate, Gareth)

Gareth did a quick check on a small number of files and confirmed that the missing neutrino triangle was gone. He is now running a full statistics validation and will advertise the results as they become available.

Are we ready to tag? (all)

The pieces we wanted to get together for this tag are:

  • FEB flasher filter - ready by the end of the week.
  • ND calib bug - fixed.
  • Spill level DQ - ready now.
  • Bad channels update (time offsets) - were in the last tag.

It was decided that we’ll wait for the FEB flasher filter - hopefully tomorrow.

Longer subruns (Jeny)

These are currently being raw2rooted. Jeny noted that they were taking a long time, with some processes waiting for > 400 mins. Discussion seemed to suggest that this was probably related to other SAM issues. She will wait until the jobs are done and then report to the group.

New job sub client experience (all)

An email thread will be started which collects our experience on the new jobsub client. This feedback will be communicated to CD.

Offsite jobs (Dominick)

Dominick began testing on offsite locations for data processing (which involves copy in and out). He had lots of errors. Nate pointed out that not all sites are setup to handle our jobs, he'll send Dominick the list of dedicated sites where things should work.

Genie problems? (all)

This is really a SIM issue. But it was noted that problems were showing up in FA releases where simulation had been frozen. This is as nu-tools wasn’t frozen...

Raw2root issues (Susan)

Jeny will reprocess the missing cosmics files all the way back to 11496 (but not the DDT).

AOB (all)

  • The FTSs are unresponsive, likely due to the computing maintenance today.