• ATLAS
• CMS
  • Lost replicas
    • When a replica is lost (last replica of a file is gone) Rucio detaches file from DS
      • Make this behaviour optional
      • Needs to be checked if this has an impact to other parts of the code
        • Rules (and derived rule states, length of datasets vs rule locks counts, etc.)
        • Locks
        • Messages
      • Should be tried out first; Adressed by test-cases
  • Process to upgrade to 1.31
    • Multihop processing in 1.31 completely in Rucio, not FTS
      • In ATLAS only done for some RSEs, not for all
  • Finished extraction of consistency checking into separate repositories
    • and pypi project
    • CMS code is now using the rucio/consistency-checking package
    • Documented under rucio/consistency-checking
      • Please send comments!
• Fermilab DUNE / ICARUS / RUBIN
  • Activated temporary table queries -> noticable performance improvement!
    • list_replicas (with and without temp tables) showed significant improvements
      • temp tables another factor 2
      • All on postgres
      • 1.29.12 was used
  • Another Rucio deployment for "mu2e"
  • Lots of work on subscriptions and automatix (for automatic testing)
• Belle II
• ESCAPE
  • Rebuilding new cluster in 1.30
  • Some problems with cs3s protocol
    • No documentation
•