LCG Management Board

Date/Time:

Tuesday 21 March 2006 at 16:00

Agenda:

http://agenda.cern.ch/fullAgenda.php?ida=a061509

Members:

http://lcg.web.cern.ch/LCG/Boards/MB/mb-members.html

 

(Version 1 - 19.3.2006)

Participants:

A.Aimar (notes), S.Belforte, I.Bird, K.Bos, D.Boutigny, N.Brook, Ph.Charpentier, I.Fisk, B.Gibbard,

 M.Lamanna, P.Mato, H.Marten, M.Mazzucato, G.Merino, G.Poulard, L.Robertson (chair), Y.Schutz , J.Shiers

Action List

https://twiki.cern.ch/twiki/bin/view/LCG/MbActionList

Next Meeting:

Tuesday 28 March 2006 from 16:00 to 1700

1.      Minutes and Matters arising (minutes)

 

1.1         Minutes of Previous Meeting

Updated minutes (version 2) of the previous meeting are available here. All changes are highlighted in blue.

 

Note: The conclusions from the latest GDB about VO Boxes will supersede, once distributed, the conclusions of the previous MB (AOB section).

 

1.2        MB in Rome update: registration

Participants should register on the Hepix site: http://hepix.caspur.it/spring2006/register.php
As explained there, Hepix registration fees are not to be paid, if one follows only the MB and the GDB meetings.

 

From the Hepix organizers:
Please note that all participants to this meeting must register at the HEPiX page for the Tuesday session, and have to pass by the main registration office and fetch their badges before entering the meeting.

 

Decision:

The MB agreed that, in order not to be in conflict with the LHCOPN meeting, the MB Meeting in Rome should start at 17:00.

1.3         Update on gLite 3.0 on the PPS (more information) – I.Bird

 

gLite 3.0 was deployed and open to users on Monday, 20 March 2006. 15 sites (5 of which Tier-1) have upgraded their PPS to gLite 3.0. The FTS topology is similar to the one that will be used in production.

 

The major work is now the deployment of SRM-based SEs. Delays are caused by bug-fixing, by the usage of GLUE 1.2 in WMS and by the fact that the version of Condor used by the WMS is not stable. But the service is available for usage.

 

Therefore experiments should prepare their tests in order to run them on the PPS. They should define in detail the configurations needed and distribute them to the working group that is checking the status of the release. This group involves members from all concerned parties (certification, PPS, development and experiments) and meets every second day. For instance, it would be useful if ATLAS and CMS repeat on the PPS their recent WMS tests.

 

Some site manager refuse to provide access, from the PPS, to their production batch and data systems. Only CERN, CNAF and PIC have opened their production facilities. All other sites are worried about compromising their production set-ups, also used by other non-LHC users. In principle there are no such risks because the worker nodes on the production service are unchanged and will continue to run LCG 2.7. The MB invites the sites to open access to their production facilities. That is the only way to execute jobs that implement realistic test cases, including MSS access and usage of the batch systems.

 

2.      Action List Review (list of actions)

 

 

No outstanding actions.

 

3.       Progress Reports for this Quarter (2006Q1) (transparencies) - A.Aimar

-            2005Q4 documents (documents)

-            Check-List for the Milestones Plans (more information)

 

 

3.1         2005Q4 Reports

The preparation of the Progress reports can (and should) be faster in the future. For 2005Q4, several reports were submitted at the beginning of February and considerable delays were caused by preparation and participation to CHEP and to the SC4 workshop in February. The process was completed at the beginning of March, when the final documents were submitted to the Overview Board.

3.2         Update of the Milestones Plans

The content of the reports was also used to update the milestone plans and feedback has been requested from the responsible of each plan. The updated plans for the sites are always available at the Site Plans page and all planning material is on the LCG Planning Wiki page.

 

In order to help the specification of new milestones a “Planning Check-list” is now available.

 

The MB agreed that high-level milestones plans should be provided also by the experiments. The initial plans will be derived (by A.Aimar) from the experiments’ presentation at the Mumbai workshop, then passed to the experiments that will first check and then maintain up to date those Milestone Plans.

 

The MB decided that collecting information from the Tier-2s is premature, and that it should be discussed again towards the end of next quarter (2006Q2).

 

3.3         2006Q1 Reports

For 2006Q1 the process will remain similar to the previous quarter, but will be done within about one month.
The proposed dates are:

-          4 April – Reports to fill are distributed by A.Aimar: Like in the previous quarter the reports will already include the milestones planned for the quarter, and the ones “delayed” from the previous quarter.

-          11 April – Sites, areas, experiments have completed the reports and sent them to A.Aimar:

-          18 April – A Review Document will be sent to the MB: The review will be done by a team (3 or 4 people) that will verify more in detail the plans contents, comment all open issues and ask for clarifications.

-          25 April - Taking into account the feedback and the requests in the review document, the reports will be verified and completed by the sites, areas and experiments and sent to A.Aimar.

-          30 April – The Review Document and the Reports are sent to the MB for approval, and then submitted to the Overview Board:

 

 

Action:

4 April 2006: The Review Team should be formed. MB Members should agree on the name of 2 or 3 people, from sites and from experiments.

 

Note: The format of the Quarterly Reports will be mostly unchanged; except that:

-          Fields for the “Resources Available (CPU, disk, tapes)” will be added.

-          Fields for “Resource Usage Accounting”, by resource and experiment, will be added

 

4.      Feedback from the Overview Board – K.Bos

 

The Overview Board discussed its role in the LCG, and agreed that it will be not only to discuss issues but also to “make decisions”.

 

The first topics discussed were:

-          Accounting:
The OB agreed that accounting is needed, should be implemented on all sites and results be reported regularly to the OB. Accounting information should be ready for the CRRB in October.

-          Association between Tier-1 and Tier-2 sites:
This association depends on the experiments’ computing models and the resources available to them at Tier-1 sites. The OB decided that by next OB meeting (June 2006) the experiments should provide at least draft documents describing their Tier-1/Tier-2 associations.

-          VO Boxes: The OB endorsed the conclusions of the February GDB, deciding that (1) experiments should not enhance their VO-specific services until the VO working group has reached its conclusions, (2) then development and middleware projects should plan the implementation of those features, and (3) that until those features are implemented the Tier-1 sites should provide VO Boxes.

In the end the VO Boxes features should be provided as general services, and therefore VO-specific implementations of these general services should not be needed further.

Tier-2 sites are free to decide whether they will provide VO Boxes or not.

 

 

Note: when the final official summary of the OB meeting is provided, that information supersedes the text above.

 

 

5.      Feedback from the LCG-LHCC Referees Meeting

 

 

Postponed to next week.

 

 

6.      AOB

 

 

No AOB.

 

7.      Summary of New Actions

 

 

 

 

The full Action List, current and past items, will be in this wiki page before next MB meeting.