Choose timezone
Your profile timezone:
LCG Service Coordination Meeting
Description
LCGSCM
Internal meeting for host laboratory LCG service providers focussing on issues on a timescale of ~1 week to ~1 month.
Mailing list: lcg-service-coordination-meeting@cern.ch
Present: J.Shiers (chairman), H.Renshall (secretary), J.Casey, G.Mccance, O.Barring, T.Kleinwort, A.Retico, S.K, U.Schwickerath, J.v.Eldik, P.Mendez, V.Boppanna, V.Singhal, M.Anjo, M.Girone, S.Campana, R.Santinelli, A.Sciaba, M.Dimou, R.Mollon, M.Littmaath
Outstanding Issues and Actions:
Reinstallation: TK- ongoing. Apel (grid accounting) problems at the moment.
Hardware for SLC4 tests: D.Smith has requested 4 boxes (ex-batch nodes- pre-agreed).
CMS Servers: HR- we should revive IT-CMS task force. I will contact CMS.
Other changes: JvE- have to upgrade firmware of 118 Transtec diskservers in February. Requires node downtime.
MG- We need to schedule expansion of RAC clusters. Should be less than 4 hours.
Highlevel milestones: As attached. JS- agreed at yesterdays MB meeting.
LCG Service Review:
Certification: As attached.
Monitoring: JC- have had 35 replies to survey of site monitoring tools.
WMS and DM: As attached. No news on the grid UI for SLC4 but the lxplus alias change to the SLC4 cluster will not be delayed.
Databases: As attached.
Auth and Auth: As attached.
LHCB: RS- see attached and FZK problem was fixed today (an NFS mount of lhcb software)
ALICE: PM- see attached and wanting to develop experiment SAM tests.
ATLAS: SC- Xmas MC production was best ever. Main problem is voms production role cannot write in user DPM directories. Have script from J-P.Baud to change ACLs and would like to test it on PPS but we will ask for long term solution of allowing secondary groups in DPM and LFC. We are also interested in SAM tests for the site vo boxes.
CMS: AS- we are testing SLC4 worker nodes. We had a problem with users disappearing from voms because their institute expired in the CERN HR database.
AOB:
TK- L.Field wants to do some bdii testing. T.Bell has discovered some older lfc and bdii nodes were installed as jbod system disks rather than mirrored.
JS- during the WLCG workshop next week please participate in the panel sessions to try and keep them lively and relevant.
There are minutes attached to this event.
Show them.