Reprocessing: Meeting of 18-Oct-2004: 9:30-10:30 ESNet video conference

Agenda

1. News
2. Status of Implementation of Reprocessing in JIM (update)
3. Detailed Timeline for Remaining Work
4. Book-keeping within JIM. (Amnon Harel)
5. AOB.

Minutes

Participants

Topics

1. News
   
   • p17 testing: 4 runs one dataset. No crashes. One infinite loop (error found, to be fixed). Reco cert plots look good at first glance. Detailed studies underway.
   • Fermilab needs to upgrade to LTS3 (Redhat Enterprise 3) possibly before the end of the year. Possible issues: JIM not yet tested for that. Install new head node (based on LTS3).
2. Status of implementation of reprocessing in JIM (update)
   Reco works. Results get stored to durable locateion. One problem with the metadata is pending. Thereafter merging can be tackled. MC merging might apply out of the box.
3. Detailed timeline for remaining work
   Metadata for merged files should take a project id from one of the parent files. Prototype should be working by the end of the month. November available for deployment and bug fixing.
4. Book-keeping within JIM. (Amnon Harel)
   

   JIM is already providing most of the information used by the LCG people. On useful table is a table of jobs which failed. That might be realised as a special view of the exisiting information. (Performance might require a special table for that).

   It might also be interesting to make the status information available that runjobs keeps in xml format.

5. AOB

Next Meeting