Suggested agenda for Collaboration meeting:

0:20  - Overview                                                                       — Maurik
0:20 - Update on data production                                            — Nathan
0:20 - DQM status                                                                   — Sebouh
0:30 - Update on tracking investigation                                   — Norman
0:30 - Update on Database & suggested improvements         — Jeremy
       => Overview of the tools, framework and documentation 

Any other suggested presentations are welcome.

Agenda:

  1. Collaboration meeting                                          — All
  2. Data Production, issues, suggested solutions      — Nathan/All
    1. New manager needed.
    2. Better manager tools?    
  3. Software Roundtable                                             -- All
  4. Questions?                                                             -- All

Software Round Table                   – All
 

Notes:

Some "software" talks may be at other places in the meeting, i.e. the tracking talk may go to the SVT session.

On data production:

The Data Production Manager title will pass from Nathan to Rafayel with the help of Kyle. 

Currently the scripts that exist do not use "swif". Nathan commented that these tools are still rather new and possibly not fully debugged. Sebouh commented that one of the advantages of using swif is the ability to have job depencencies, but such dependencies don't really exist for the reconstruction. 

Nathan also mentioned that for a previous data processing setup he used to write the status of the job to text files. In the end he found that you always need to check the files that were supposed to be produced, so the advantage of using these text files is limited. Nathan does have a script to check the stub files to see what jobs need to be re-run. He will let the current pass6 jobs finish and will then asses how to continue.

Software management:

The run up, and false starts, to pass6 shows that we are not doing great with our software management. We briefly discussed this, but there does not seem to be an obvious way to do better. We should think about better ways to manage the releases of software and the testing of a release, to make future production passes less painful.

New people to take over tasks:

We are starting to see turn over of key people in the experiment. We need to start thinking about who can take over some of the key tasks. One of the big ones that comes to mind is GBL and alignment, which were in Pelle's domain and are not fully completed.

 

 

 

 

  • No labels