SamWiseMinutes4 14 14¶
Chris, Dominik, Andrew, Michael, Nate, Ryan
Adam – analysis workshop for MINOS?
Eric — ? Status of FD GENIE —
- Run some more MC to exercise GM01 and GM03 —
— FD GENIE is most important - (ND GENIE if FD is finished) — just need overlay specific metadata.
— ND overlay is low priority – testing overlay is higher priority. (need online tag for multipoint)
- Need calibration and tag to start reconstruction: would be great to run some jobs by 3pm meeting tomorrow.
- data Raw2root jobs if available – wait to start keep-up processing after some data has been reconstructed.
CAF files: Andrew – good idea to register CAF data.
- Doing CAF right would be useful to get files to user’s that would be in a more final format.
- Is there enough metadata there to register the files?
- Dominick will follow up.
- reading the constants from the database blows up the memory usage (all on the database side).
—> Need to get Jon involved…
Dominik — waiting on calibrations
Matthew — Got constants from Chris B. last week.
—> Luke has some test jobs running.
—> Some progress expected early this week.
—> Thought there was something related to the SAM meltdown.
— log files missing – check for .log file…
—> Luke’s jobs failed from image size. Memory usage.
Can we understand the difference in apparent performance for the two SAM stations:
01: SL5 (~250 per hour) — MC files
02: SL6 (1000 per hour) — Raw2root files.
Can we stand up another VM like 02? — Craig submitted a ticket.
Over the weekend we witnessed superior performance from samgpvm02 (SL6) compared to samgpvm01 (SL5) - however they were transferring very different types of datasets. We would like to request a 3rd VM samgpvm03 (SL6). We would like it to have configured like samgpvm01 but an SL6 OS. In addition, we would like the the directories that are currently being watched by sampvm01 to be split between 01 and 03. In this way, we can check the performance of the two VMs under the same kind of dataset environment.
From:kanika (not available for the meeting today):
The only action item related to me was to make a new tag with version 3 calibration as default. Version 3 of calibration is not ready yet, the absolute calibration processing is in progress. Once that is complete, I'll make a tag.
Of other note on software side are two things:
1) Robert Hatcher provided me some scripts from MINOS that log what releases are being sourced. I've adapted them to nova and we are now logging this information. Soon I will review what releases are commonly used and move towards deleting old, unused releases from the VM's to make more room.
2) we are nearly ready to update art on offline. New version of artdaq compatible with new version of art is imminent this week. I'll begin testing novasoft with new art once that is available. Before I move offline forward to the new version of art, I will ask for 'go-ahead' from production and samwise groups. Before I test everything, it's hard to say how disruptive moving to the new version of art will be. I'll keep those concerned updated on developments on this front.