<!-- keep this as a security measure:
* Set ALLOWTOPICCHANGE =
TWikiAdminGroup,Main.LCGAdminGroup
* Set ALLOWTOPICRENAME =
TWikiAdminGroup,Main.LCGAdminGroup
#uncomment this if you want the page only be viewable by the internal people
#* Set ALLOWTOPICVIEW =
TWikiAdminGroup,Main.LCGAdminGroup,Main.ChippComputingBoardGroup
-->
CHIPP-CSCS Face to Face Meeting on 2018-06-21
- Date and time: Thursday 21st of June at 10:15
- Place: Zurich Zentrum (LEE E 126 link)
- External link / EVO: Only if requested
Agenda
- 10:15 - Welcome, coffee and agenda
- 10:30 - VO Status report (~last 6 months + changes since last F2F, for both Phoenix and CRAY)
- LHCb (20' - Roland)
- ATLAS (20' - Gianfranco)
- CMS (20' - Thomas)
- 11:30 - Tier-2 status, plans & pledges
- CSCS (45' - Various people, for both Phoenix and CRAY)
- UNIBE-LHEP (30' - Gianfranco)
- 13:00 - Lunch
- 14:15 - Tier-3 status and plans
- PSI (15' - Nina)
- UNIBE-ID (15' - Gianfranco)
- UNIGE (15' - Gianfranco)
- 15:00 - Coffee break
- 15:15 - NGI_CH (20' - Gianfranco)
- 15:35 - Tier-0 activities (25' - Pablo & Gianfranco)
- 16:00 - Discussion (30')
- 16:30 - End of meeting
Attendants
- CSCS:
- CMS: Christoph Grab
- ATLAS: Gianfranco Sciacca
- LHCb: Roland Bernet
Minutes
## LHCb
- Site is okay and things are working fine.
- Higher failure rates on Piz Daint, not correlated with the type of job, which should be followed up.
- The problem is that Vladimir does not follow up (problem disappears and he does not care anymore). Roland shall speak with Vladimir to try and see what to do.
## ATLAS
- Site looks fine, comparable with LHEP
- arc04 is behaving badly towards SAM tests (jobs work well, though). It will be followed up between ATLAS and CSCS
- Failed WC is high, which needs to be checked (it might be already solved).
- We should have a separate meeting regarding putting ATLAS dCache pools under NGDF (lead by Gianfranco, with Stefano and Pablo).
- Regarding Tapes, Gianfranco will lead the dicsussion with ATLAS but CSCS would like to be involved in the technical discussion (e.g. use Swift instead of S3).
## CMS
- CSCS went down in the classification quite significantly
- Resolve times are good (fast, once start working on the issues) but it takes much longer before that.
- Accounting monitoring issues are still present and makes it difficult to judge how much is being delivered. CSCS plots will show this more precisely.
## CSCS
- Pledge page split for each VO should include % next time (to help see the fair-share effect),
- CSCS should check if Slurm is killing jobs for excessive time, this might explain why LHCb has that many failed pilots
- Pablo to send a doodle for next
F2F in January
- Worth checking with Pascuale, who is already running on
UserLab for LHC on GPUs, who also has experience in running in
JupyterHub
- Follow-up on open tickets in GGUS 132927 and 133787.
- For Tier-0: CMS may face limitation on the firewall on the SWITCH side, good to know in case of trouble
Other Action items
Attachments