PASTE meeting

Tuesday 12 Dec 2006, 10:00 → 12:00 Europe/Zurich

Andrei Tsaregorodtsev

CERN, bat 2 R-030 To join the telephone conference, click here: https://audioconf.cern.ch/call/0122882 To dial in to the conference: a. Dial +41 22 7676000 b. Enter access code 0122882

Production and PASTE Meeting, 12 December 2006, CERN Chair: A.Tsaregorodtsev Minutes: S. Paterson CERN: G.Castellani, P.Charpentier, J.Closier, S.Paterson, R.Santinelli Phone: N.Brook, C.Cioffi, A.Tsaregorodtsev Apologies: M.Bargiotti 1. Production Status Report ( J.Closier ) Joel reported that the requested MC productions are running smoothly. The output of Boole is being directed to Tier-1 SE tape storage and the output of Brunel is stored on Tier-1 SE disk. Replicas of the DST should be stored on all Tier-1 SE disk but for some previous productions this was not the case and should be resolved ( ACTION: Joel ). Philippe pointed out that so far we have produced ~1M events for one signal so far. However, to provide the opportunity for feedback on the latest Brunel release, a limited number of events (~50-100K) should be produced for all channels. Workflows should be prepared for this ( ACTION: Joel ). 2. Operations Debugging continues for reconstruction jobs at NIKHEF, recent failures were due to SRM failures. Stuart commented that a fix is in CVS to catch a related error. gsidcap protocol should be tested at RAL, GridKa and IN2P3 to confirm that the current mechanism (dcap:gsidcap) works in all cases ( ACTION: Joel ). Joel also reported that when rescheduling jobs sent to IN2P3, the destination site DIRAC.Lyon.fr is automatically set. Stuart mentioned the same behaviour has been observed for jobs sent to CERN and DIRAC.LHCBONLINE.ch. The rescheduling logic should be checked ( ACTION: Andrei ). Andrei mentioned that the software installation area at Lyon had become corrupted, this should be investigated in install_project.py. Andrei also reported that the functionality to blacklist a site is now present via the production tools / console. Roberto has developed a tool to streamline the query of SRM for TURLs. This is not yet fully tested but in the future can be used to automatically determine available protocols from the LCG IS. Stuart reported that the RB protection mechanism is in place on the test system, this will be part of the next DIRAC release. 3. gLite RB Usage ( G. Castellani ) Gianluca reported that testing of the gLite WMS has uncovered some bugs which have been entered into Savannah. Contrary to LCG job submission, it is necessary to specify a Wall Clock Time requirement with gLite jobs. Philippe and Nick were not aware of this and will discuss this at the next Grid management board ( ACTION: Phillippe, Nick ). A new variable (PLATFORM) to distinguish between submission to glite or lcg will be introduced to the Agent Director service ( ACTION: Stuart ). The gLite Agent Director module will then be merged with the current Agent Director ( ACTION: Gianluca ). 4. Data Integrity Checks ( A.Smith ) Joel confirmed that all data before DC04 can now be scrapped. Andrew reported that a tool has been developed to poll the LFC for zero size files. In the event a zero size LFN is found this is reentered into the catalogue and resolved. This has been tested for DC06 files (DIGI and SIM). Andrew also mentioned that the same logic can be used to correct zero size replicas. The plan is to run this in a weekly cron job ( ACTION: Andrew ) 5. AMGA BK Status ( C. Cioffi ) Carmine has updated the BK webpage with the latest layout for the interface. A BK-LFC integrity check is still to be added ( ACTION: Carmine ). Eventually will require a Data Management console operation to consistently remove a file from the LFC and the BK FC interface. 6. DIRAC3 Brainstorming Session ( A.Tsaregorodtsev ) The Agenda for the DIRAC3 Brainstorming session next week is now available. This will consist of a series of half day sessions. For each session there will be a convener who will introduce briefly the current situation in the given DIRAC subsystem, the strong and weak points, open questions. This is needed to start the discussions. The conveners are also supposed to take notes and prepare a summary by the last concluding session. ( ACTION: Ricardo, Stuart, Andrew, Gennady, Joel, Andrei ) 7. AOB Conveners for next week's DIRAC3 Agenda are requested to contact Andrei to confirm their availability.

10:00 → 10:15 Production status report

- Report from the Production meeting:

10:15 → 10:45 Operations

- Software installation issues - Production tools enhancement - NIKHEF tests and related issues - gLite RB usage

10:45 → 11:00 gLite RB usage

- Status of the gLite director agent - Use of the PPS for gLite/SLC4 tests

Speaker: Gianluca Castellani

11:00 → 11:15 Data integrity checks

Speaker: Marianne Bargiotti

11:15 → 11:30 AMGA BK status

Speaker: Carmine Cioffi

11:30 → 11:40 Agenda for the next week DIRAC3 brainstorm session

11:40 → 11:50 AOB