PPS Pilot Follow-up Meeting Minutes Thu 02 Sep 2008

  • Date: Thu 02 Sep 2008
  • Agenda: 38527
  • Description: pilot of Cream CE: check-point
  • Chair: Antonio Retico

Attendance

  • PPS: Antonio Retico
  • CERN: Harry Renshall
  • CMS: absent
  • Alice: Patricia Mendez
  • Atlas: not represented
  • LHCb: not represented
  • CNAF: Daniele Cesini
  • FZK: Angela Poschlad
  • JRA1/Cream/WMS: Massimo Sgaravatto, Alessio Gianelle

Status and results of the pilot service (by VOs and sites)

Feedback Alice (Patricia)

Patricia presented the progresses done in the last weeks. She showed a presentation from the last Alice Task Force meeting (28-Aug-08)

The Cream CE, is showing a stable performance after it was introduced in production and no special attentions or interventions were required during the summer

The next step is to point the Cream CE toward the Alice production queue. Alice will do it as next thing

Alice is not using at all (nor plans to use in the future) the WMS-based submission to Cream. In support to the direct submission they had to require the set-up of a gridftp server at the site. This was provided by Angela at FZK in to the VOBOX.

The gridftp server is needed, within the direct submission, in order to retrieve the output sandboxes without having them temporarily stored on the Cream CE, which could cause the service node to collapse.

The gridftp server was removed from the glite3.1 implementation of the VOBOX on demand of the LHC experiments but the Alice is going to require it to be re-introduced.

Alice asked the developers for a status update on the new upcoming version of Cream (CE). Eventually the experiment, seen the results, is willing to go in production with the current version.

Alessio: is not using the WMS a global decision for Alice? Patricia: It is, as soon as cream will be in production the WMS based submission will be used only as a fall-back solution

Site Admins feedback

No further feedback (apart from what mentioned by Alice) was given by the site admins

Status of Development/Deployment of future versions of Cream

Massimo: PATCH:1755 was certified and went in PPS, it was supposed to go to production last Wednesday. A patch to fix the WMS in order not to match the cream CE was release well before. So there is no risk for new WMS to match Cream. This patch was not released for SL3. Cream should publish something different from "production".

This is tracked in BUG:40705, fixed by PATCH:2201 now in certification. As soon as certified, the current version of Cream can be moved to production.

Antonio: there will be however a short round in PPS for the configuration chages applied by this patch. Are the other bugs fixed by PATCH:2201 to be tested in preproduction?

Massimo: They are only fixes in YAIM and all of them are quick to test or transparent

Antonio: Applying this patch to the services already running in Karlsrhue does not satisfy our test case and it is not in the scope of the pilot. We may do it to keep the services synchronised but we still need to do a specific test on top on PATCH:1755 We will request one pre-production site, possibly Upatras, to join the pilot immediately after the deployment test and let CMS do a quick test (it will have to be WMS based)?

Patricia: Alice is not impacted, however by a change in the production state advertised by the CE

Antonio moved the question whether the priority should be raised toward at EMT for the certification of PATCH:2201. Alice does not see it as a priority for production, although they are eager to see the Cream CE deployed. As the likely timeline for certification of this patch is however estimated (by Antonio) to be of the order of 1 week, the decision was made not to request an increas3 of priority to SA3.

Alice would like to have an idea of the delivery date and the version. In order to start the negotiation with the sites for the installation of the Cream CE

Massimo: The version to use is the one shipped with PATCH:1755. It works fine for direct submission.

Antonio: if we don't set a specific priority on PATCH:2201 it could be certified within one week

Alice: However Alice is taking the risk if something goes wrong with the new service. We will keep however theLCG-CEs as fall-back solutions

Alessio (about the status of the WMS PATCH:1841): no official news but it could be ready for certification this week or at the beginning of next week

Antonio: so, adding another 2 weeks for certification. in about three weeks we could be ready to take this WMS in PPS and start phase2 . We will ask then FZK and CNAf to re-install from scratch the WMSs.

Open Issues (by VOs, sites, deployment teams)

List of Open bugs and relevant decisions

Recommendations for release and deployment

Decision about termination/extension of the pilot

In consideration of what discussed above, the decision is made to extend this phase of the pilot within the PPS infrastructure of extra 4 weeks.

The next check-point meeting will be on the 30th of September

AOB

Daniele will step over TASK:7142

Actions (Review of tasks)

Status of the subtasks of TASK:7143 .

Some changes done off-line

Notes:

* TASK:7159 gstat was updated and can now work with Cream CE> The task is closed * TASK:7274 , TASK:7279 : The service management tasks were updated as a consequence of the decision to extend the pilot (see later)


These minutes can only be changed by members of:

Edit | Attach | Watch | Print version | History: r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r2 - 2008-09-04 - AntonioRetico
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    LCG All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright & 2008-2019 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback