CESER LIGO / UCSD UChicago call

Tuesday 13 Aug 2019, 13:00 → 14:00 US/Central

http://bit.ly/rwg-zoom

Edgar Fajardo Hernandez (UCSD), Frank Wuerthwein (UCSD), Igor Sfiligoi (University of California San Diego), James Clark (Georgia Institute of Technology), John Hicks (Internet2), Pascal Paschos (University of Chicago), Robert Gardner (University of Chicago)

Google drive folder, https://drive.google.com/drive/folders/1OaDZ6OPE6On8Bt2byuoAWr5kKH7XwmUZ

Meeting notes in:

https://docs.google.com/document/d/146OP2zqQC7BB4SAmgLuv-TeYLTTesiPs_NfCttdwliA/edit#heading=h.i2ed2c4usz4v

2019.08.13 UC-UCSD OSG-LIGO CESER

 

Minutes are also saved in indico  by the end of the call.

 

Attendees:James, Edgar, Pascal, John Hicks

1. John Hicks updates

Still on a holding partern. The Amsterdam machine is in Ann Arbor. Other machines are still on hold. Delivery to I2 by mid-september. Texas TACC? Original argument is to get coverage in south. LIGO does not have an HPC XSEDE allocation. Can we replicate CMS/ATLAS on Stampede2 for LIGO. Use a stratum for LIGO software/data? Edgar: not for data. 

 

Hicks: a. Google drive for I2. People will need to start populate it. 

B. We are targetting the site - Not the cache. So a user can not designate it cvmfs. 

C. Set up meeting with Pascal. 

2. osg.Ceser? Gratia probes were getting in but since the project was not created. This is issue is now resolved and we see the rsv jobs reporting there. 

3. RSV to ES injesting. Capture the Critical/NonCritical. Display it a time series per site (walltime / cpu time). We can not account for the reasons for issues. John can correlate with network. Can we modify the messages for RSV to differentiate between time outs or errors?