Current ATLAS items:

• ATLAS will first run carroussel retrievals, 1% data flow will start next week once carroussel tests are over.
• GC configuration: Julien to circulate current numbers.
• EOS rate limiter configuration is not persistent, however it is found in a script. Julien will call it from the EOS startup script.
• Nodes not belonging to production setups should be reassigned to a different host group (e.g. spare). This is the case eg for eosctaatlaspps-srv-m1. (action on Julien)
• Virtual drives and tape pools should be removed from production setup (Julien)
• Others:
  • Monitoring items to be developed:
    • file and volume counters per tape pool
    • pending and late file archive and retrieval requests (important for assessing system health)
  • eos info <file> still returns misleading information (such as /does_not_exist and fsid=65535). This requires to be fixed with the EOS team (important, but not urgent)

Actions follow-up:

• Storage classes: Pending answer from Alessandro.
• Web page: Done by Julien, will be moved to main Git branch
• Activities and FIFO: Still ongoing. Eric to circulate updated timelines.
• ATLAS deletions: Julien confirms ATLAS will invoke "eos rm" via Rucio under normal circumstances.
• New hardware: Bernd confirms hardware will arrive in September, but will then be subject to testing and installation. High priority to talk to procurement and network people (to ensure all network infrastructure is in place when nodes arrive) (Julien).