Project

General

Profile

4-January-2016

Attending: Satish, Alex, Prabhjot, Sijith, Tapasi, Chris, Qiulan, Justin, Gavin, Vito,

News (Satish/Alex)

We will soon be implementing an ECL instance for production. The idea is that we will use it not as a replacement for the production mailing list, but as a way to detail exact commands used, output and status updates, and refer to ECL entries to document exactly what was done so we can go back for reference.

SW Tags (Paul S)

The lats release Paul made was the Chewbacca release (S15-12-07). Kanika also wants to make a release for MRE, but isn’t quite ready to do so. Paul is planning to make feature branch for new version of art (v1.18). This is in anticipation of the move to root v6, and we anticipate that there will be some pain in making the switch. The feature branch is an effort to ease the pain of that somewhat. There was some discussion of how long this feature branch should last, given that we don’t want to be making maintaining both this and development for a long time. It was suggested that we merge the feature branch back into development after all the required fixes are made (provided they are backwards compatible). Migration to root v6 is not anticipated until after the summer conferences.

There are some upcoming minor changes in art v1.17.04. They are not relevant to us, but we should update to the newest version when convenient to stay current. This will also require a nutools version bump. There are some issues with external dependencies that also need to be resolved to get the most recent version of nutools (currently v1.18) working for us.

Alex also provided an update an the status of github. We have a github repository and the CVN group is testing it out. Testing for novasoft has not begun yet.

Rationalizing MC Generation (Paul R/Gavin)

As a first step Paul is planning to move the submission scripts to NovaGridUtils. Paul is also spending some time learning the details of the software. Paul will work on rationalizing the softer tools overall. Gavin is working to rationalize the command line options. We’ve set a deadline of having this ready for the collaboration meeting.

Processing Status

Reco Keepup (Paola)

Paola was not present to provide an update, but Vito presented what he is aware of. These jobs have been proceeding smoothly with no big issues. Seven files have failed with exit code 255. It was not clear in the meeting if this correlates to any well known error. Vito commented that these are often transient errors and may be recovered with the next round of processing.

Raw2root Keepup (Vito)

Four files are failing, and they have been reported on the ticket. One file exhibited a known failure, and Satish marked it as bad. The other three files failed because of large memory consumption, and probably need to be marked as bad also. Satish is following up with the DQ group.

Horn off data (Enrique)

Of the 2,806 artdaq files, 2330 have been processed all the way through. The remaining files cannot be processed because of excessive memory consumption. They are all DDActivity trigger files. Only ~800 of the caf files have shown up in SAM, as the FTS rules for CAF only accept the comics and numi files. Chris suggested that we generalize the FTS rules so that all files are accepted. Satish volunteered to do this. Update On investigation this looks more complicated than initially expected. Satish will send an email to the list.

Miniprod ND Genie (Bruno)

Bruno is on vacation, and hence had no update.

Miniprod FD Genie (Tapasi)

One file is missing in the gain 100 artdaq dataset because of some confusion in the draining dataset. Tapasi will address this. There was also some confusion because of six nonswap gain 100 files that appear to have come directly from the FCL file rather than from the g4 file as they should have. We may simply need to get rid of these files and not try to regenerate them. However we should still attempt to understand what went wrong. Tapasi and Satish will investigate. Flux swap and tau files are complete.

AOB

Prabhjot asked about the status of the six missing FD CRY Epoch3b pclist files? The last information suggests that the files were produced but did not make it into SAM. A service desk ticket was filed, but there has been no update. Alex will try to track down the issue.