AFS weekly meeting

Monday 15 Jul 2024, 11:00 → 12:00 Europe/Zurich

600/R-002 (CERN)

AFS service/operation meeting

11:00 → 11:20 AFS ops

AFS service/operation meeting

hardware failure
how to avoid involving the cloud team
Arne's input, warm standby + VLDB update
Jose's input and the new cloud resources

cold recovery tests, updated statement
https://cernbox.cern.ch/external/eos/project/i/itsd/IT-SD-Group-Resources/BCDR/Cold%20Recovery%20Plans/SD-GSS.md?app=CodiMD

RQF2630026 perf qual metric
RQF2626035, ticket routing

Fons absence

Decisions:

• Investigate the "warm standby" scenario which involves reattaching rbd volumes and updating VLDB
• Follow up on how we could do the VM "evacuation" if necessary (Arne proposed a rundeck job)
• Old harware
  • Plan a move, after understanding if "warm standby" increases our options
    
    • if not, normail drain workflow
• Cold recovery tests
  • Write and exercise a procedure to create a new AFS cell and restore a single volume there
    
    • Need to find relevant backups on CTA
    • Our restore machinery presently assumes the CERN.CH cell
• Follow up on the Service Perf Qualification stuff
  
  • https://it-td-service-performance-qualification.docs.cern.ch/IT-SD/Services/afs/
  • RQF2630026