Status of the Production ALICE TF MEETING 25/02/2010
Full production this week Workflow for this week Most of the activity of this week comes from user analysis jobs Full production this week User jobs this week Pass1 reconstruction activities going on: Pass 5 reconstruction activities at T1 sites finished at the end of the last week New MC (production) cycles still to be agreed by the Physics Groups Following Latchezar’s explanations last week, massive MC production is not expected before data-taking
Activities and issues for this week Central services Air conditioning failed at the AliEn central services server room 19 Feb at 04:00AM Central services unavailable also affecting the DAQ team Sites IN2P3-CC: The « verylong » queue announced during the last ALICE TF meeting has entered in production this week Good behaviour of the new service GRIF: CREAM-CE available for ALICE. However the information provider of the system is publishing some inconsistencies. This is preventing the system to enter in the production environment of Alice for its testing
Activities and Issues for this week CASTORALICE @ T0 DAQ observed several timeouts while transferring data from PIT to CASTOR@CC yesterday after lunch Problems from xrootd or disk servers were excluded Issue seems to be solved at around 14:00 (no actions applied) Issue followed by the CASTOR experts, which have enabled the trace for eventual future issues GLUE Schema discussion next week Triggered by ISS (Adrian Savcenco) The aim is to understand from the experts how to publish information related to any local services (in particular any CE) into the IS We have prepared a phone conference with Laurence Field (IT-GD): Tuesday (2 March) at 10:00 Call: +41 (0) 22 76 77 000 Phone Conference: Glue Schema setup for ALICE sites Responsible: Patricia Mendez If any site is willing to join, please do not hesitate to contact on Tuesday
Migration of SAM to Nagios CERN is migrating the current SAM infrastructure to Nagios One of the new requirements the system has is that all services to monitor must be properly registered in the gocdb/is This is preventing the monitoring of ALICE VOBOXES in 22 sites We were able to pass over this requirement with the previous SAM infrastructure that we created in the pass In this way, each time a new site/service was entering in production we had to register it in just one place Meeting with the experts is foreseen for this week We will try to figure out if we can avoid this requirement for the VOBOXES If this is not the case, sites will have to ensure their VOBOXES are properly registered in the GOCDB/IS
List of affected sites INFN-CATANIA BOLOGNA-T2 ITEP CERN INFN-CAGLIARI KOSICE INFN-TORINO IPNL INFN-LEGNARO SARA IHEP PRAGUE CNAF MEPHI BUDAPEST GRIF_IPNO CCIN2P3 SINGAPORE GRIF_IRFU FZK INFN-BARI GSI