PASTE meeting

Tuesday 7 Sept 2010, 09:30 → 13:15 Europe/Zurich

2-R-014 (CERN)

Andrei Tsaregorodtsev

CERN, bat 2 R-014 To join the telephone conference, click here: https://audioconf.cern.ch/call/0165202 To dial in to the conference: a. Dial +41 22 7676000 b. Enter access code 0165202

PASTE Meeting, 7th September 2010, CERN

Chair:  P.Charpentier
Minutes:  F.Stagni

CERN: Z.Mathe, F.Stagni, S.Paterson, A.C.Smith, V.Romanovskiy, M.D.Rashid, R.Santinelli, P.Charpentier, J.Closier

Phone: R.Graciani, A.Zhelezov

0. Development tasks:
  0.1. Bookeeping
  For the last week we've no progress on the merging productions, and stopped the ongoing reconstruction. RAL problems also stopped us from completing. 90% of the jobs failed when checking for descendents in the bookkeeping. A lot of calls to the getDescendents() method causing the service to stuck. Zoltan says that most of the calls come volhcb20, and he's suggesting a possible bug with the DataRecovery agent, Andrew thinks that it's not possible. Anyway Stuart stopped for 10 minutes the DataRecovery agent to see if it is causing the excessive load, but this does not seem to be. Zoltan will have a second look. Another experiment is to restart the bookkeping service to see if the number of threads still increase: this seems to be true, because after 5 minutes we already had 27 threads running. It could be a wild loop in bookkeeping code.

  0.2. RSS activation
  We decided to activate a real ban (and un-ban) of the sites and SEs, for the following cases:
    - an OUTAGE downtime on the site is announced: 12 hours before it starts, the site in banned
    - an unscheduled OUTAGE downtime on the site starts
    - when all the CEs of a site are in OUTAGE downtime, the site is banned (12 hours in advance, if possible)
    - when the SRM interface is OUTAGE donwtime, the corresponding SEs are banned, just when the actual downtime starts.
    - when a SE has 0 space left, then it is banned for writing.

  Some basic management will be put on CS side.

  It will be investigated the possibility to use the notification system for eLogg-ing these actions.

  0.3 Throttling
  Job throttling is done and tested, will be in the next release.

1. Production Issues
CondDB problem. We now persist the information in the CS. Information about it should be forwarded to the lhcb-grid ml so that GEOC could update that.

2. DIRAC Release Update
Some fix may require a patch release.
For what regards new releases that should pass the certification tests, nothing too urgent, so we'll wait for Andrei to be back for that.

3. DIRAC Tasks
Nothing major to report.

4. AOB
NTR
 

09:30 → 10:00 Ongoing actions

The task list is available via the Savannah DIRAC group ( Task List ). If you are not already registered please do so at the Savannah page. In order to streamline the time spent during PASTE on the actions list only tasks in the status 'ToBeDiscussed' will be discussed there. In case some actions fall by the wayside they will automatically be updated to this status. Therefore the action assignees should take care to update this. For ongoing actions it is advantageous to report any progress in the Savannah tasks each week.

10:00 → 10:30 Production issues

- Site throttling

10:30 → 10:50 DIRAC Release Update

- next release - Certification test tools

10:50 → 11:10 Data caching tests using DIRAC FC

11:10 → 11:40 DIRAC tasks

- Infrastructure monitoring (SAM,SLS,Dashboard) - ResourceStatus System status - Framework status - Bookkeeping status - WMS status - DMS status - Production Management status - VO-boxes - Web Portal - Services deployment ( hardware, configuration ) - Documentation

11:40 → 11:50 AOB