AFS weekly meeting

Monday 25 Jan 2016, 14:00 → 15:00 Europe/Zurich

31/1-012 (CERN)

AFS service/operation meeting

AFS operations meeting 2016-01-25

present: Dan, Massimo, Jan

Issues:

• big AFS outage last week (14h, OTG0027970), service got deleted from AD (hence no tokens), Lessons learned?
  • will stay on "best effort" (after all, was working for AFS..)
  • SLS probe needs to go "red" next time, currently is unauthenticated ( AFS-376 )
• AFS phaseout:
  • will get contact persons from LHC experiements. Who else?
    • BE -> A.Bland, contact next week
    • TH -> E.Gianolio?
  • Scanned for 7001/UDP, stats are on JIRA, will try to redo to track over time (if OK with security team).
    • per-version/OS Summary data: should go into Grafana
  • ATLAS says "EOS+FUSE is slow" (quality monitoring histograms, web service). want more details..
  • OPAL data has been moved form CASTOR to EOS - could ask to do the same thing for software area (into EOS project), and if OK, phase out their AFS space? (Massimo to contact M.Schroeder once all data transfer done/validated)
  • Discussion: could create & administer EOS project space via OpenStack MANILA, with EOS-specific plugin, could replace "afs_admin". Check overlap with FIM / EGroups creation, account lifecycle workflow. If easy, might even hook up AFS resources (readonly, for display / migration tracking?)
• one ATLAS volume cannot be backed up since December (tried salvage and 'vos move'). Perhaps Kuba will have a look once back..