02/15/2013 Minutes


Jan, Susan, Nick, Craig, Jarek, Louise, Jon, Ji, Minerba, Chris, Kirk, Himansu, Raphael, Mat, Gavin, Nate


Nate: MC generation in S12-11-16 almost complete for increased ND stats, as submitted by Nate. Increased from 140 files (1k events each) to 14000 files, or 1.4M ND events.
As if this meeting there were 248 files still to complete processing. They are all finished now. They were initially submitted Wednesday 13th so completed in approximately 48 hours.

Action Item: Nate will reconstruct these files in S12-12-12.

Jan: We are almost ready for tagging today. See Gavin's doc-db 8677 for details of what has been added and needs to be added.
We did tag 2 weeks ago and required minimal backports. We now have fixed NovaDatabase to HEAD version in Offline.
All Online packages have been brought up to their HEAD versions, including DAQDataFormats - requested by Jarek for Nearline processing.

Action Item: Jan will tag Saturday, S13-02-16 (DONE)

Action Item: Gavin to commit a new module that will read in a MCTruth object for cosmics and spit out a single MCTruth for each cosmic interaction. (DONE)

Mat noted that the File Transfer System (FTS) was in a state of flux and was failing on multiple moves to SAM.
Note that this can be monitored using the following:

Action Item: Follow up with Robert Illingsworth about the FTS and the dropbox area.


There was a Metadata workshop run by SCD that we started to define the metadata we would like to store in the files.
Marc Mengel pointed out that he spoke to Chris Green with regards to putting metadata into ART files and was directed to the following page:

Specifically this indicates one should be able to do:

ServiceHandle<art::FileCatalogMetadata>()->addMetadata(key, value);

with string key,value items pretty much anywhere in your Art code, and the key/value pair ought to show up in the output. Adam successfully was able to put metadata into an ART file as well as extract this.
Action Item: Define what Metadata we want stored in the files.


We had a poductive session with Marc Mengel talking specifically about how best to write our production scripts.
In the end we decided that we should archive our generated fcl files to SAM and then SAM would allow one to create a dataset from fcls for generation.
Mat has kindly agreed to write the fcl generator or more specifically a Fcl Basemaker and a Fcl Jobifier.

Action Item: Mat to produce Fcl Basemaker/Jobifier in time for next full MC generation round - circa several months.

QEEventFinder Validation

Nick presented doc-db 8680 which showed validation in different detector configurations of the new QEEventFinder "QePid" PID module. He requested this be able to the next round of processing. This request was granted.

Action Item: Add QePid to S13-02-16 PID production. (DONE)


Next production meeting is Friday March 1st 2pm.