Report from the Executive Director

September 8^th

This is the first report since the Council face-to-face in August. Activities of the OSG project of note are:

1. Regular Engage Users use of OSG resources of >25,000 CPU hours/day. While this represents 5-10% of the overall throughput it must be noted that the use is limited by the users need, not the available cycles. When the usage is >40,000 CPUhours/day the job queues become deep. We are starting to look into whether making more resources available to the VO helps with this.

2. Finishing the campaign to help sites offer D0 opportunistic use of storage, and improve the end-to-end efficiency/cpu load of the application.

From Joel Snow: DZero MC OSG production rose 2.1M events to 9.4M events in the week ended September 8, 2008. This is 127% above the average weekly OSG production rate for the past year of 4.1M events. Overall Efficiencies improved. OSG sites with SE's efficiencies: MWT2-IU 0.81, UCSD T2 0.55, UNL 0.67, MSU 0.67. 3 OSG sites with eff. <10%:

Also the CPU utilization/load (cpuhours compared to Wallclock time) is regularly – but not consistenly - improved.

We believe that significant progress has been made, an improved plateau of overall effectiveness reached, and that the next steps should wait for the data management task force internal to D0 to executive its plan. OSG is putting together a report as a result of this week.

3. The Grid Operations Center has upgraded the BDII service which is critical to US ATLAS and US CMS running. We are now running scalability and robustness tests. Following that a high availability service will be implemented. The first phase is scheduled to be in place for the October LHC data taking run.

4. Torre has been informed that the following paper has been accepted: M anuscript ID: ci-2008-00263c Title: "An Open Science Grid study of the coupling between conformation and water content in the interior of a protein" Author(s): Damjanovic, Ana; Miller, Benjamin; Wenaus, Torre; Maksimovic, Petar; Garcia-Moreno, Bertrand; Brooks, Bernard We are pleased to inform you that your manuscript has been accepted for publication in the Journal of Chemical Information and Modeling. Your manuscript has been forwarded to the ACS Publications office.

5. We received an announcement from SciDAC? that the two papers submitted for the July SciDAC? PI meeting are published as part of volume 125 of IOP Publishing's open access Journal of Physics: Conference Series. http://herald.iop.org/jpcs_vol125/m66/hxp//link/1910

6. Migration of >60% of the OSG sites to OSG V1.0 release.

7. VDT focus on support for WLCG middleware client tools needed by US ATLAS and US CMS. VDT Storage focus on xrootd/Bestman to understand the individual components, testing and release structure. ATLAS expects to have 5-7 Tier-3s deploying this technology over the next few months.

8. Success publishing of OSG US ATLAS and US CMS Tier-2 RSV/OIM data was achieved through the WLCG Sam/Gridview services. OSG now publishes monthly reports to the Joint Oversight Team of the accounting and availability/reliability of the US LHC Tier-2s and Tier-1s - equivalent to the WLCG paper reports. We are starting to gather initial information about the US LHC Tier-3s.

9. First quarterly meeting with SBGrid which exposed several areas of instability in the end-to-end system and a concern that the ongoing effort needed to maintain a working OSG site is ~0.5 FTE rather than the 0.1FTE that they anticipated. These issues have delayed the expansion of the SBGrid community infrastructure. We are putting in place a task-force to look into the technical issues and to maintain ongoing contact and support.

10. Completion of the draft FY09 work plan for review by the Executive Team (sub-group meets on 9/13) such that SOWs can be in place for the planned 9/30 date.

11. Putting the new organizational infrastructure in place including receiving the needs from CMS and ATLAS for FY09.

We have been told by the JOT that OSG will have a SciDAC? continuation review in January or early February. We know little more than this at the moment.

We would like to request Stakeholders to input needs for the Communication and plan. The FY08 plan is attached. What is the best way to handle this?

I was asked to contact EGI for how OSG might collaborate in any way. I spoke with Dieter Kranzlmueller who has been the head of the EGI-DS project. Dieter suggested a "letter of support" or interest in collaborating which should be endorsed as from a National Grid Infrastructure. Talking with Bob Jones and Erwin Laure when we were at CERN a couple of weeks ago leads us to believe there are transitions in progress (Dieter moved from Austria to a university in Germany.) There will be an EGI day at the EGEE '08 conference at the end of September, and I would propose to wait for any next steps until a few weeks after that.

-- RuthPordes - 08 Sep 2008