Matthew, Dominick, Satish, Bernali, Bruno, Nate, Jeny, Paola, Ryan, Joseph, Jonathan, Chris, Susan
- There was an email from Peter S. about odd runs where all trigger streams had been merged together, currently we have no tools to split this file back up yet. We will need to find someone for this task.
- We've run out of disk space for the novasoft account which has a quota on /nova/app/ - this is where slf6 builds goes. In addition /grid/fermiapp, where the slf5 build goes is shared among many experiments and is 96% full. Jonathan will work to address this urgently, finding a short term solution today to get a new tag out then working with computing to get a medium/long term solution.
Nate is working on generating three samples at the moment: the hadronic model variations for simulation systematics; the ND PoT weighted subrun; and the post-shutdown FD genie. He had some issues which meant things got off to a slow start but currently has just about completed the generation and declaration of the 20k fcl files needed for the ND PoT weighted subrun. After some discussion with Ryan it was decided to give priority to the hadronic simulation samples. The easiest way for Nate to do these is to re-generate a baseline sample at the same time. The other samples can then follow later (likely post collaboration meeting).
FA post-shutdown MC reco (Joseph):¶
Joseph has one more test to do then will likely be ready to submit for real.
Keep-up reco (Jeny)¶
Jeny started these jobs running on Friday. To date all the ND has been processed, as has all the FD (although ~3% of these jobs failed). Preliminary studies indicate that these files are promising. Jeny will chase up the remaining failures and make datasets for wider circulation. It was noted that the "restrictedcaf" dataset are restricted access - i.e. the unblinded files. We will not circulate datasets with these files in at this time, only the blinded ones.
Keep-up calibration (Paola)¶
The validation files requested by calibration have been completed by Paola and passed over to them for evaluation.
FTS version (Dominick)¶
Dominick wants to update the FTS version to get some fancy new features. As Nate is the only other person using FTSs from production at the moment, and he is using a separate instance, Dominick will go ahead and do this now.
There was some discussion about Moving all FTS machines to bare metal machines rather than VMs. Currently we can't compare performance between our one real machine and the VMs as they scan different sets of files. If we changed one over however we could do this. Matthew will discuss with Robert I.
Genie simulation filter changes (Ryan, Susan)¶
We filter out events at genie generation that don’t make any fls hits, this affects PoT accounting. No one seems to think that it is necessary to do this for genie files so the decision was made to remove it in the make_sim_fcl logic. Nate and Ryan will do this.
migrating runNovaSAM to NovaGridUtils (Satish)¶
Dominick implemented some improvements only visible in new tags that should be useful to all users. Satish wants to migrate runNovaSAM to NovaGridUtils so all users get the cutting edge version (possible as all releases use current NovaGridUtils). This was deemed to be a good idea.
Collaboration meeting plans (Matthew et al.)¶
The plan is to have Dominick talk through the components of the production system and then Satish give a SAM submission tutorial. This will be followed by Jonathan talking about Jenkins, then a quick normal production meeting. Matthew will advertise the tutorial.
- Dominick is in progress of being committing the first merged pre-LEM job.
- It was decided to wait until after the collaboration meeting before starting a new version of rapid-turn around as if we started it now it wouldn't be ready in time meeting anyway.
- Unless an email indicates otherwise there will not be a meeting on Monday January 26th.