AFS weekly meeting

Monday 24 Apr 2017, 14:00 → 15:00 Europe/Zurich

31/1-012 (CERN)

AFS service/operation meeting

AFS operations meeting 2017-04-24

(meeting cancelled, report is FYI)

• DB server upstream update:
  • AFSDB4 (wigner slave) first updated to "upstream" 1.6.20.1, then reinstalled into CC7
    • tested on new testcell-db-cc7, looked OK, but missed some changes in "production" and for non-testcell setups..
    • 3ware controller claims to not be supported on CC7 (seems to work OK, though). SNOW ticket to see whether this can be changed.
    • ABS broken on CC7 (bad dep. on old SLC5 python-26, complains about "md5" ). For  now only affects DB servers (which do not have volumes that are worth backing up), but will have FS soon.
  • WIP: AFSDB2 reinstalled with CC7 + 1.6.20.1. Doing manual Puppet runs etc..
• VM FS servers: quota OK'ed by Bernd, waiting for OK for "xlarge" VM type.
• Some recent "puppet_wrong" errors seem to be caused by a single puppet run waiting forever (or until killed after 24h).Bug, or we launch something that gets stuck
• More "Selinux" errors (in particualr on DB servers). Ignoring for now (since we run in "permissive" mode).
• Usual work.boinc restarts, but again one case where our monitoring did not catch this (perhaps linked to nightly backups? all stuck..). Question from Riccardo whether they could get powers to restart (not for now).
• several changes to get rid of /afs/usr/local
• several tickets (access to old files etc). 1 "very dissatisfied" user.. 1 open ticket from Rainer on "su" +CC7 behaviour..