(no real meeting, just notes from the corridor)
- Dan handled several support tickets 2 weeks ago
- Kuba had one "overload" case eventually traced to the use of CMT inside the LSF job (which creates a symlink forest and causes heavy stat()). Asked ATLAS for help on how to run such (Athene) jobs efficiently
- accepted 1 new project (aware of migration but claims HTCondor and EOSFUSE don't mix yet).
- accepted 1 quota extension (+1TB) - beams thing, move to EOS planned in LS2
- ABS bug: should wait for tape recall to finish (after stager_get), else subsequent "xrdcp" might time out (user-visible error)
- ABS will be pointed to CTA in ~2 weeks, expect one major round of changes (and pending cleanups) in ~4 weeks
- Next ITUM will have a note that anything around migrating /user, /work is on hold pending the new EOSFUSE version
- AFS servers in Wigner (triggered by oldest open AFS SNOW ticket...)
- setserverprefs script seems to have no effect (or is not run) on CC7? I.e. CC7 Meyrin clients will use the (2) fileservers there
- puppet CellServDB does not mention the DB servers there - i.e these are utterly unused (and useless)
There are minutes attached to this event.
Show them.
