AFS weekly meeting

Monday 15 Dec 2014, 14:00 → 15:00 Europe/Zurich

31/1-012 (CERN)

AFS service/operation meeting

AFS operations meeting 2014-12-15

present: Massimo, Kuba, Dan, Jan

Issues:

• "best effort" schedule for the Xmas break - see SNOW calendar for the "critical" services incl AFS
• CDS/INSPIRE/INDICO: 2 old "critical power" servers reinstalled (after some back-and-forth), announced to T.Baron; still moving volumes. AFS-294
• 1.6.11pre1 - wanted to test on testcell, but no RPMs yet (easy enough to recompile but need then to host yet another YUM repo). Discussion - could be on linuxsoft, but not really worth to create a new repo just for this..
• Q: can we use a different cell to "park" volumes  (for disaster recovery)? not really, would be better to investigate "shadow" volumes at Wigner (at least within same cell..)
• ABS: CASTOR nameserver issue has appeared again, apparently after "our" machine became again part of the general DNS alias. Agreement Kuba/Xavi:watch for 2 days, decide whether to go for own CASTOR NS again.
• ABS: TSM restore was broken for ~6 weeks, only one user noticed (and was right at the boundary between TSM and CASTOR backups..)
  • general: can start to phase out TSM restores in January.
• mdmam: do not have the spare "chassis" configured, i.e. auto-rebuld won;t happen for most RAIDs. Dan will fix this..
• vendor accounts (itprocos) need "sudo" access (otherwise risk of ping-ponging between sysadmin and repair team).
  • might be controlled via Hiera (sysadmins_exclude_vendoraccounts), pulls in "sssd".
  • anyway will need some "sudo" right for individual sysadmins/operator accounts in the future