Ceph/CVMFS/Filer Service Meeting

600/R-001 (CERN)



Show room on map
    • 14:00 14:05
      CVMFS 5m
      Speaker: Enrico Bocchi (CERN)
    • 14:05 14:10
      Ceph Upstream News 5m

      Releases, Tickets, Testing, Board, ...

      Speaker: Dan van der Ster (CERN)
    • 14:10 14:15
      Ceph Backends & Block Storage 5m

      Cluster upgrades, capacity changes, rebalancing, ...
      News from OpenStack block storage.

      Speaker: Theofilos Mouratidis (National and Kapodistrian University of Athens (GR))


      • ceph/dwight issue over weekend: cascading OSD failures on p05151113613837, seems like the kernel went bananas (nothing logged anywhere since 10pm saturday). Solved by rebooting the host.


      • Reformatted EC02 on ceph/erin
      • We still view the unnecessary HEALTH_ERR due to backfill miscalculations
      • The error gets away when a recalculation happens (it may take quite a time)
      • Each rack takes about 3-4 days for the whole procedure to finish
      • Racks EC03 to EC06 are next.
      • It is estimated that in two weeks that the reformatting will be finished.
    • 14:15 14:20
      Ceph Disk Management 5m

      OSD Replacements, Liaison with CF, Failure Predictions

      Speaker: Julien Collet (CERN)


      • Handled a ticket that affected a host in the ceph/erin's EC02 pool that was about to be reformatted at the time.
      • The disk replacement script produced the following message:
        • /dev/sdac has no OSD mapped to it
      • Should this be treated as an error? The device in not used by ceph, therefore the replacement procedure through ceph scripts may be skipped
      • I informed the guy to report again if he encounters this error and not take any action
    • 14:20 14:25
      S3 5m

      Ops, Use-cases (backup, DB), ...

      Speakers: Julien Collet (CERN), Roberto Valverde Cameselle (Universidad de Oviedo (ES))

      Dan and Giuliano

      • New nomad configuration:
        • Dedicated rgw's for gitlab and cbox
        • CVMFS will go through default rgw's from now on
        • Current status: 3 rgws each for atlas, cbox and gitlab, 4 for defaulT
      • Traefik configuration changes:
        • Each rgw has now 2 traefik frontends (CEPH-779)
          • 1x host-based routing
          • 1x pathprefix-based routing
        • Traefik keepalive feature disabled (CEPH-778)
      • S3 Accounting:
        • Prototype version works
          • Collect user info -> map user to dep/group -> push output file to S3
        • Needs to coordinate with Hugo on the desired output format and to be cronified
    • 14:25 14:30
      CephFS/HPC/FILER/Manila 5m

      Filer Migration, CephFS/Manila, HPC status and plans.

      Speakers: Dan van der Ster (CERN), Pablo Llopis Sanmillan (CERN)
    • 14:30 14:35
      HyperConverged 5m
      Speakers: Jose Castro Leon (CERN), Julien Collet (CERN), Roberto Valverde Cameselle (Universidad de Oviedo (ES))
    • 14:35 14:40
      Monitoring 5m
    • 14:40 14:45
      AOB 5m