SCDailyMeetingTemplate (edit)

Week of 061016

Open Actions from last week: bdii107 and 8 firewall changes will be made Monday after which they can be put in production, one to exp-bdii and the other to lcg-bdii (via FIO load balancing). Move atlas WMS UI from rb101 to rb103 Monday morning.

Chair: H.Renshall

Gmod: A.Retico

Smod: U.Schwickerath

Monday:

Log: Atlas moved to rb103 in afs wms UI

New Actions: Put bdii into production.

Discussion: CERN SRM timeouts Saturday due to limited number of connections (a config file was missed - corrected Saturday evening). ALICE moving servers today/tomorrow.

Tuesday:

Log:

New Actions: Change bdii load balancing from 2 of 3 to 3 of 3 (TK). Migrate myproxy-fts to px102 and 104 (ML,HRR).

Discussion:

Wednesday

Log:

New Actions: No time yesterday to move myproxy-fts so do it today (done at 14.30). Put rb109 back in production.

Discussion: ce107 is not scheduling jobs. The rbs do not 'see' it. To be followed up (HRR). CMS export is running well. Following a problem at FZK they were able to catch up at the increased rate of 140 MB/s. rb109 has been returned by Elonex.

Thursday

Log: myproxy-fts was moved to an HA solution and rb109 was put back in production, dedicated to CSA06.

New Actions: Install rb110 (ex lxfsrb5901) for CSA06 use (return rb101 to GSSDATLAS when done).

Discussion: ce107 is still not scheduling jobs at CERN. It is however visible as a CE at DESY. CMS data export will drop for a while as they finish a sample but then increase up to double the current rate. GSSDATLAS will stop their T0 exercise and data export tomorrow then resume export after a few days. CERN needs to recuperate the Atlas scratch tape pool. Failure of a RAC server again seemed to cause the FTS (and perhaps PheDex) database to lock up. A fix is due to be applied shortly.

Friday

Log: glite-wms UI files in afs were replaced yesterday by version which fixes libtar 111-character filename bug.

New Actions: Install rb110 as a gridrb and dedicate it to CSA06. Continue investigation of ce107 non-scheduling of jobs.

Discussion: GSSDATLAS data export has stopped for now as planned. CMS export stopped from 02.00 CET, initially thought to be a PheDex problem but later found to be another DB lockup. Fix is in PPS but already validated so it was agreed to deploy it on Monday.

Your signature to copy/paste:

Force new revision help | | or or or Access keys: S = Save, Q = Quiet save, K = Checkpoint, P = Preview, C = Cancel