Attendees

Gouaty, Livio, Gergely, Antonella, Nicolas, Didier, Benoit, Alberto. Cristiano, Scott Koranda, Adam Mercer, Patrick Brady, MAB, ....

Agenda

Pending action items

- S6 and VSR2/3 RDS data transfer to CC (Livio et al)

- datafind at CNAF (MAB, Gergely)

- Data migration Castor --> GEMSS at CNAF (Karen C. / Antonella)

- virgo_gpfs3 disk increase (Karen C. / Antonella)

- Test the access to data transfered by DT at CNAF (Gergely) and CCIN2P3 (MAB)

Projects

- Pilot job project status (Romain/Gergely)

- VDT: data access at CNAF and Lyon (Gergely/MAB)

Minutes

- S6 and VSR2/3 RDS data transfer to CC:

Scott has sent the L1 missing files at Cascina. Livio is doing the transfer at CNAF. Should be finished tonight. Check the .tmp files (MAB already checked the files are present and OK). Livio is also transfering VSR3 RDS data transfer.

- Datafind at CNAF (MAB, Gergely)

Files generated by MAB are not accessible by Gergely (home protected). Files are now accessible.

- Data migration Castor --> GEMSS at CNAF (Karen C. / Antonella)

- virgo_gpfs3 disk increase (Karen C. / Antonella)

Karen asked a date. Alberto will coordinate that with the CW group.

- Test the access to data transfered by DT at CNAF (Gergely) and CCIN2P3 (MAB)

No news from Geregly and MAB. Next time

Projects

- Pilot job project status (Romain/Gergely) slides

Geregly: condor flocking has been used to be able to launch jobs at CNAF from any LSC cluster. Tests done all OK. Still one pending issues: most of the cbc pipelines assumed that all disk areas are shared. That may be a problem if part of a full workflow is run in different places (CNAF and somewhere else).

MAB: is that a pb for cbc follow-up pipeline?

Romain: a prori not.

Gergely mentions he has been able to run a full analysis pipleine there. What's next?

Everything is ready for the cbc pipelines to run at CNAF. Who will use it?

1. cbc follow-up pipeline (need to access Virgo rawdata). In Romain's slides tehre is the list of the LSC sites that have enabled condor flocking jobs.
2. Romain is proposing to set up a utility tool to let anyone to run omega scans at CNAF
3. Patrick Brady explains that Milwaukee will test running jobs 

UWM: 2 submit hosts (trout, marlin)
CIT: 2 submit hosts (pcdev1, pcdev3)
Syracuse: 1 submit host (sugar-dev1)

That should be the priority ordering. Patrick Brady also mentions the long term strategy: an end user should be able to run everywhere without knowing/bothering where to run. The system would balance according to the cmputing availability.
MAB: agreement on the long term strategy. The short term strategy is to have end users testing and using CNAF in order to discover pbs.
Gergely: monitoring web pages give indication of what is run at CNAF, but does not report job errors.

Action items for MAB, Romain and Gergely:

1. make pilot job working for VSR3 data
2. follow-up what is run at CNAF.
   

 
Next meeting: wed dec 1st.
                     agenda: follow up of pilot jobs + VDT + data access of files transfered