Swiss Grid Operations Meeting on 2020-12-10 at 14:00
Minutes:
CSCS report (the only one uploaded)
- WLCH SAM: 100% availability and reliability
- CHIPP pledge delivery: 111.3 %
- shared as 41% ATLAS ; 33% CMS ; 26% LHCb
- Reminder: during xmas break Piz Daint maintenance is on best effort - no monitoring
- ARC CE state counts is now in the dashboard (Nov 25)
- ARC CE Log Access:
- Hardware received, configuration underway
- All three experiments need to explicitly agree to share this data between each other and this needs to be on record
- Nick wants the confirmation that by making the log available were not breaking any confidentiality agreement
- (ATLAS/CMS members cannot have certificates for both experiments for confidentiality reasons)
- --> Gianfranco / Derek is that OK ?
Gianfranco: agreement for ATLAS logs is implicit for the operation team. Derek and Roland are part of it, nothing explicit needed. ATLAS will review possible log ownership and access issues in due course, since ETHZ is not an ATLAS institute. E.G. we might need an explicit agreement in order to run an ARC CE for ATLAS in a non ATLAs institute. FYI: VO certificates have nothing to do with "confidentiality". They are part of an Authn & Authz infrastructure covering all VOs on the Grid. Like I don't use electricity from my neyghbour or if I play with BSC YB I cannot simultaneously play for FC Lugano. Nothing to do with confidentiality.
- CMS dCache Access:
- Access to dCache data via Macaroons
- Waiting on some feedbac
- > Derek
- ATLAS DPM Migration:
- Convert part of dCache to DPM
- Test in progress, need plan for making production
- > Gianfranco
Gianfranco: see ATLAS VO report
- Validation of accounting data:
- Khours vs KHS06 Hours Accounting Discrepancy:
- BDII and ARC configuration seem correct
- Requested assistance from Julia, no response
- Ticket in the process of being opened to request assistance
- > Mauro fwd Dmitrios contact
- Overlapping ARC CEs
- One Slurm partition served by 2 ARC CEs
- If this is an OK configuration, then go ahead during the February 10th maintenance
- > Gianfranco: Can we get an example configuration?
Gianfranco: already explained in the relevant discussion that on the user side, configuration is OK. If not, we'll iterate. Please post arc.conf in slack if help is needed.
- this would improve the 2 ARCs unbalance we have today
NEXT MEETING 21 JANUARY 2021 @ 14h00 - Let me know if that is not OK for you
Followup from previous Action Items
Action items
VO reports
ATLAS
- ATLAST2ReportNov02020.pdf: ATLAS CH Tier2 report
Highlight: ATLAS CPU usage at CSCS has sharply declined and is getting worse. The site will eventually be phased out as Nucleus and this can have a severe inpact on the computing strategy for ATLAS in Switzerland. CHIPP and CSCS are invited to appoint in the immediate term one person to deal with this long-ongoing issue, in order to resolve it with dedication and my assistance.
CMS
LHCb
T2 Sites reports
CSCS
T3 Sites reports
PSI
EGI / WLCG
Review of open tickets
a.o.b
- Attendants
- CSCS: Nicholas Cardo, Pablo, Elia
- CMS: Mauro
- ATLAS:
- LHCb: Roland Bernet
- EGI: