LCG Management Board


Tuesday 16 January 2007 at 16:00




(Version 1 - 18.1.2007)


A.Aimar (notes), D.Barberis, N.Brook, F.Carminati, T.Cass, Ph.Charpentier, B.Gibbard, F.Hernandez, M.Kasemann, H.Marten, G.Merino, B.Panzer, , Di Qing , L.Robertson (chair), J.Shiers, O.Smirnova, R.Tafirout

Action List

Next Meeting:

Tuesday 30 January 2007 - 16:00-17:00 - Phone Meeting

1.      Minutes and Matters arising (minutes)


1.1         Minutes of Previous Meeting

No comments. Minutes approved.


2.      Action List Review (list of actions)

Actions that are late are highlighted in RED.


  • 19 December - J.Shiers and H.Renshall will report on the progress on the definition of targets and milestones for 2007 at the LCG ECM meeting.


On the way. Discussed during this meeting.


  • 19 Dec 2006 - The proposal to ALICE is to consider, as in the TDR, a value of 10**6 for the ALICE ion runs. L.Betev agreed that ALICE should confirm it within a week.


An email message from Y.Schutz reopened this issue (Y.Schutz’s email, D.Barberis’ email).

Will be followed-up outside the MB. F.Carminati said that ALICE is checking the values and will report to the MB.


  • 15 Jan 2007 - D.Foster will form a group discussing the network setup and performance needed according to the Megatable values (for T1-T1 links on the OPN, etc).


To do.


Additional information to be provided on accounting:

         IN2P3 – status of automated reporting of non-grid CPU.


Distributed an email with the information required (see email)

In summary, IN2P3 is currently not reporting non-grid jobs into the GOC DB. If they are requested to do it the can.

But they would like to be able to distinguish what is grid and what is non-grid when they submit their data.


         NDGF – status of automated reporting of CPU usage (grid and non-grid); plans for user level accounting.


The information is available for grid and non-grid usage in their system, but it is not interfaced and published in the GOC DB.

O.Smirnova confirmed that their goal is to be ready to publish the information by March.
Note: NDGF cannot distinguish for non-grid usage which VO it belongs to (not even if is a WLCG-related activity). Therefore NDGF should then report only grid usage.


         OSG – plan for the OSG accounting system to report CPU accounting data automatically to the APEL repository; plans for user level accounting


R.Pordes could not be present at the MB. She distributed an email with status and plans of OSG accounting (see email). Was just received before the MB meeting; therefore it is better that she will summarizes the status of OSG accounting at next MB meeting.



Other points from Last Meeting

         Ian Bird – to report on the deployment status of the scheduling priority scheme information providers


I.Bird distributed an email reporting that the release deployment would be available soon. L.Robertson asked further clarifications about whether the sites should proceed on mapping experiments group and roles onto their systems or wait for the release.


M.Schulz sent an email just before the MB meeting (see email) suggesting that sites should go ahead while the deployment team also install and certifies the patches on the PPS.


         All Tier-1 sites – to plan as a matter of priority the deployment of the VOMS groups and roles priorities scheme; sites not using Torque to organize integration with their batch system

A date should be fixed for the sites to study and implement their solutions and report to the MB their initial progress.


Proposed date: Tuesday 13 February 2007.


  • 31 Jan 2007 - Sites should send to H.Renshall their procurement plans.



3.      GSSD Working Group: Mandate, Membership, Chair


K.Bos could not be present.


F.Carminati confirmed that the mandate distributed by K.Bos in December and again in January (see email) did not receive requests for changes. Therefore he asked for the MB’s approval of the GSSD mandate.


Decision: The MB approved the GSSD Working Group mandate.


The MB agreed also on the proposed chair: F.Donno will chair the group.

The attendance is still not complete and the WG itself should decide on the appropriate attendance (see email).


N.Brook noted that the attendance is too high and will be difficult to organize such large group.



25 Jan 2007 - F.Donno should report to the MB (by email) the list of people that will participate to the GSSD working group.


4.      Update on Targets and Milestones for 2007 (SRM 2.2 critical methods) – J.Shiers


4.1         Proposed Milestones

1. Demonstrate Tier0-Tier1 data export at 65% of full nominal rates per site using experiment-driven transfers

-          The load generators for ALICE, ATLAS and CMS should be ready by the time the tests start.


2. Demonstrate Tier0-Tier1 data export at 50% of full nominal rates (as above) in conjunction with T1-T1 / T1-T2 transfers

-          The proposal is to use the inter-Tier transfer targets taken from ATLAS DDM tests and from CMS CSA06 targets. Something equivalent will have to be defined for ALICE and LHCb.


3. Demonstrate Tier0-Tier1 data export at 35% of full nominal rates (as above) in conjunction with T1-T1 / T1-T2 transfers and reprocessing / analysis stripping at Tier1s

-          Tier1 production activity needs to be defined per-VO and depends on computing model and production plans of each VO.


4. Provide SRM v2.2 endpoint(s) that implement(s) all methods defined in SRM v2.2 MoU, all critical methods pass tests

-          Proposal that requirement on bring online methods be relaxed as that they are not required in Q1 2007 (i.e. can still pass milestone without them, e.g. success with).
SRM 2.2 critical methods are attached, experiments were asked to comment at the ECM meeting. That list was agreed and is a subset of the SRM 2.2 MoU.

4.2         Questions

With respect to milestone 3 above, could a generic sub-milestone be written that is VO-independent, e.g. every file transferred to a Tier1 is read at least once by a Grid job running at that site?

-          This would allow VO-specific milestones, e.g. that couple to conditions DB deployment models, to remain as experiment milestones, which is probably where they belong.

-          Some explicit targets for the WMS at each Tier1 would nevertheless have to be established (derived from the above).


Does it make sense to have (an) explicit database milestone(s) at this level? (Which would clearly have to be experiment-specific).

-          The same comment as above applies. Trying to keep it experiment-specific and make sure that all data is read and written to the DB at least once.


The MB agreed on the milestones above. At the WLCG workshop J.Shiers will present them as milestones for 2007Q1.


D.Barberis reminded that ATLAS needs the “bringOnline” method. J.Shiers agreed that this will be available, but he is not sure that will be ready for Q1.


L.Robertson reported that he has asked M.Litmaath to present at the workshop a high-level schedule for the MSS deployment and mapping to the real sites storage systems. The BOF on this topic is on Thursday 25 January. In the second quarter the tests will require the SRM 2.2 functionality. Therefore must be deployed on the Tier-1 sites.


L.Robertson noted that must also become clear how the SRM functionalities will be deployed and tested (by real experiments use cases? or F.Donno’s test suite?)


5.      Reliability and Availability Data Summary for December 2006


The reliability data summary has been distributed via email. The latest version is here:


The target was just met in December with “8 best sites average” above 88%. Sites should send again a report about their reliability and availability.



22 January – All Tier-1 sites should send their site report on December 2006 reliability.


A summary, like G.Merino did, will have to be prepared for mid-February (in order to also cover January 2007).

If nobody volunteers, L.Robertson will ask some Tier-1 MB members to do the summary.


R.Tafirout asked on how many past months the averages should be calculated.

The MB agreed that the average should now be split in: (1) “last 4 months average” and (2) a “global average” from the beginning of data collection.


6.      Follow-up on Groups, Roles and Job Priorities


A.Aimar reminded the sites that J.Templon had offered help from SARA on the implementation of job priorities.
But also that it had been agreed the sites should start with the implementations on their batch systems (LSF, Torque, BQS, etc) without waiting further.


H.Marten asked, and all MB members supported the request, for documentation from SARA on how to implement the priorities, examples and any existing useful information.


F.Hernandez reported that IN2P3 has implemented an agent that modifies the priorities of the jobs in BQS. Was used already by ATLAS in November 2006 in order to guarantee that at least 80% of the executed jobs at IN2P3 were ATLAS production jobs.

IN2P3 will then see how to implement the priorities in BQS itself; but the solution used for ATLAS could be used for the time being.



22 Jan 2007 - J.Templon should distribute pointers to all information and documentation that can help sites to implement job priorities on their batch systems.


7.      AOB



No meeting next week.

Next MB meeting on the 30th January.


8.      Summary of New Actions




22 January – All Tier-1 sites should send their site report on December 2006 reliability.



22 Jan 2007 - J.Templon should distribute pointers to all information and documentation that can help sites to implement job priorities on their batch systems.



25 Jan 2007 - F.Donno should report to the MB (by email) the list of people that will participate to the GSSD working group.



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