Notes kindly taken by Craig Group, edited by M.Tamsett.
Here are my notes from today:
Craig, Gavin, Chris, Tom, Dominic, Robert I, Eric, Nate, Paula, Jeny, Kanika, Matthew, Alex, ...
SAM database migration (Robert Illingworth).¶
FNAL will migrate the the Backend database from Oracle to PostgreSQL (for financial reasons). This must be done by January for next Oracle update. In addition some improvements will come along with this migration. From a user point of view it is mostly the same, however there will be some minor changes, for instance, the date presentation format will change. Also, multiple checksums can be allowed and the checksum file will look different. At the same time there will no longer support for multiple projects, which is a historical feature that no-one seems to be using anyway.
It will be important for us to verify that PostgresSQL can perform at the level we need. The colloquial evidence seems to suggest that perform should be comparable overall and possibly even more stable. A risk identified is that if we need to migrate back to the old system after a few days then there will be lots of data to move back to Oracle.
Action Item: Dominick is at Fermilab so he will have a sit down with Robert this week to discuss stress tests.
News from the simulation group about process towards the first ana tag (Nate)¶
Simulation met on monday & dedicated to what should go into the tag. They would like to get Adam’s updated noise simulation in for sure. They have also been updating the ND geometry based on Susan's studies and measurements by Xuebing & Nate. So far Xuebing & Nate have performed measurements on one side of the ND. They intend to measure the other side in the near future.
Action Item: What about the FD alternative geometry? Craig will follow up with Jim.
Unified setup scripts (Kanika, Gavin)¶
They've modified the setup_novasoft to setup the nova environment including a bunch of hooks to catch failure modes. At the same time they've refactored the standard script and the one for CVMFS machines into a single unified script.
Progress update on raw2root keep up (Jeny, Paula)¶
They met with Dominick and have some issues need to meet with an expert. They were optimistic about being able to take this over in a week or so.
Feedback to Michael about the dashboard website (All)¶
Beta SAM monitoring website. However Michael was not present.
Action Item: Matthew has collected feedback from the group and will meet with Michael on Tuesday to discuss this.
Genie cry overlays (Nate)¶
He has some files ready for the Sim group to validate, they ran pretty fast. Once validated we'll move onto producing a large sample.
Dominick points out that FTS01 seems like it has a backlog. Nate will move the 01 directories so that they are “watched” by 02 and 03. In a related note Robert has new version of FTS — v4 will be ready soon.
It was pointed out that SCD will deploy a new jobsub client in October. Forcing the migration of the jobsub client in October. Eric is the only one who tested it so far. There will be a release at the end of August or early September.
Action Item: Craig will get us invited to FIFE meetings in future.
Action Item: Matthew et al. Come up with a plan for testing this before deployment.
ART stakeholders meeting this afternoon.¶
Gavin & Kanika will attend and point out that the changes to metadata required for art were a big problem for us. We need the new release.
- Kanika will make a tag for Gavin later today…
- Gavin is speaking at g-2 workshop tomorrow on behalf of the production group.
- Jon pointed out that there is a bug in the validity database web server. Can’t upload data that overwrites old data. This needs some downtime to update the web server? If it goes well this won’t take long, and should not affect our jobs. We don’t have too much going on now, so we would like to go ahead and get this done.