LT2 Technical Support Group

Monday 26 Nov 2007, 14:30 → 16:00 Europe/Zurich

532 Blackett Lab (Imperial College London)

David Colling

The meeting will be reachable by phone conference at 0845 862 0333 pin: 963852

Present: * IC: D.Colling, M. Aggarwal * IC-LESC: K. Sephton * UCL: G. Sciacca, Ben Waugh * QMUL: G.Mazza * RHUL/BRUNEL: Duncan Rand * ATLAS : Lucio Apologies: * LHCB: Raja [status via email] 1. Site Status * IC-HEP & IC-LESC & IC-ICT -- MA: Ongoing problem with ATLAS jobs failing at ICT site. DR: Let us wait for the new PANDA job submission for the site. -- MA: Atlas prd accounts problem fixed for the LeSC site. -- Keith: New prd/sgm accounts will be added soon on the Mars cluster. * UCL-HEP & UCL-CCC -- GS: UCL-HEP -- (no problems) -- UCL-CCC -- In downtime for the new cluster installation. William will start the testing of the new cluster from tomorrow. The current plan of the cluster testing is: Till-Mid-Jan (Local testing), Mid-dec-software installation testing and lcg job submission testing. Duncan will submit a script to do lcg testing. David suggested for bulk testing using new glite-wms. They will also perform 24hr throughput test along with data transfer from the SE. * QMUL -- GM: Problem with SE last Friday. Rebooted the machine. However the load seems to be very high on the SE. Mostly due to access to mysql database. Long term solution will be new SE. -- DC: QMUL SAM tests performance is very low. -- GM: More information required for SGE cluster running no shared home directories. -- GM: For the new cluster installation -- looking at the new SGE jobmanager suggested by Keith. * RHUL & BRUNEL -- DR: RHUL - OK. There are many biomed jobs running on the cluster but they are not doing much. -- DR: BRUNEL -- CMS production jobs submitted by Maiko running out of SE disk space. * AOB --- DR: FTS transfer problems from UCL & QMUL (Check the DPM verision installed). 2. Experiment problems/issues (If there any) * ATLAS -- Lucio (Atlas job Efficency per site) QMUL -- 90% RHUL -- 90% BRUNEL -- 83% IC-HEP -- CE- cx1 ( jobs failing ongoing problem) UCL-CCC -- down UCL-HEP -- job proxy expired error. Jobs are staying in the queue for a long time. Set the high priority for ATLAS jobs. * CMS -- DC: Not much going on. DR: Received no response to his email to Pete regarding CMSSW installs in London * LHCb -- Raja: LHCb has nothing special running at the moment on the Tier-2s. * Others: -- Supernemo : DC: They have used lt2 resources as ltwo (VO) for nemo analysis. 3. Open tickets & Action List -- Tickets: 2131 --- PPS ticket 2173 --- Ongoing 2205 --- (Remove LFC from the site BDII) -- UCL (Close the ticket) 2206 --- (Remove LFC from the site BDII) -- UCL (Close the ticket) 2218 --- Fixed. Testing ongoing. 2225 --- Fixed. -- All Actions: Open ALL: Install Supernemo (VO) in LT2 (CE's and SE's) GF: Check the version of DPM installed at UCL. GM: Check the version of DPM installed at QMUL. DC: Update Olivier's information on GOC database etc. DC,DR,MA: set up LFC for ltwo and supernemo (VO).

14:30 → 14:35 Site Status

* IC-HEP & IC-LeSC * UCL-HEP && UCL-CC * QMUL * RHUL * BRUNEL

14:35 → 14:40 Open tickets & Action List

14:40 → 14:45 Experiment problems/issues (If there any)

Atlas CMS LHCB Others