Indico has been upgraded to version 3.1. Details in the SSB

Rucio Meeting

Europe/Zurich
Martin Barisits (CERN)
Videoconference
Rucio Development Meeting
Zoom Meeting ID
413496641
Host
Martin Barisits
Alternative host
Mario Lassnig
Passcode
28849311
Useful links
Join via phone
Zoom URL
    • 3:00 PM 3:05 PM
      News 5m
      • Rucio 1.27.0rc1 feature release candidate out
        • Identified a bug breaking DIRAC
          • https://github.com/DIRACGrid/DIRAC/issues/5637
          • DIRAC runs without rucio.cfg, one workflow in 1.27 requires it
            • Should be easy fix, will build an RC2
        • Identified a second bug in the transfer-submitter
          • Will also be fixed for RC2
        • Currently running on ATLAS integration
        • Plan is to have 1.27.0 final next week
      • 1.27 retrospective next week (Dec 2)
        • What was achieved? Which milestones did we miss?
        • How can the planning be improved?
      • 2022 Release schedule planning (Dec 9)
      • 1.28 Release roadmap planning (Dec 16)
      • December meeting schedule
        • Dec 02
        • Dec 09
        • Dec 16
        • First meeting in 2022: Jan 13
    • 3:05 PM 3:15 PM
      Community News & DevOps roundtable 10m
      • ATLAS
        • 1.27.0rc1 deployed on integration
        • Analytics nodes / dumps export
        • Submitter mixup with activities
          • Causing troubles with MH transfers
      • CMS
      • ESCAPE
        • DAC21 ongoing
        • Inconsistency with rule creation for replicas
          • ~5% (of the data created in the last month) of files have no rule associated
            • Trying to reproduce it, but could not for now
            • Asked LSST who are the majority owner of these files
              • Using a simple `rucio upload` for these files - which should work fine
              • Ask LSST to use --verbose and we will have a look into the logs
              • Possibly upload of an entire directory will not cause rule creation?
        • OIDC client dependency #5031
          • 4 daemons will crash if there is an issue with OIDC dependency
          • Only load clients at the first time a token will be used
        • Restart policy (24h restart policy)
          • Needed due to thread crashes
        • Reports by prometheus do not correspond
          • Mislabeling of plot?
          • Followup with CERN K8s people
      • SKAO
        • Noticed daemons (conveyor submitter) stop writing output after a few weeks
          • Implement restart policy
    • 3:15 PM 3:25 PM
      Hot topics 10m
    • 3:25 PM 3:55 PM
      Developers roundtable 30m
    • 3:55 PM 4:00 PM
      AOB 5m