r4 - 05 Mar 2011 - 10:14:35 - RichardMountYou are here: TWiki >  AtlasSoftware Web > Minutes28Jan2011

Minutes28Jan2011 RAC Minutes, January 28, 2011

Members (*=present, #=apologies)

*Richard Mount, Kevin Black (Physics Forum Chair), *Jim Cochran (Analysis Support Manager), Alexei Klimentov (ATLAS ADC), Ian Hinchliffe (Physics Advisor), Rik Yoshida (Tier3 Coordinator), *Michael Ernst (U.S. Facilities Manager), Rob Gardner (Integration Coordinator), *Kaushik De, (U.S. Operations Manager), Armen Vartapetian (U.S. Operations Deputy)

Ex-Officio: *Torre Wenaus, *Stephane Willocq, Mike Tuts, Howard Gordon

Approval or Correction of Minutes

On hold since Richard's BNL account had expired and no minutes were posted yet.

Operations Report (Kaushik)

Many changes were underway.

At the Tier 2s, the migration of all data from MCDisk to DataDisk was in its final stages. As of two hours ago MCDisk had been made readonly.

At the Tier 1, the migration was proceeding more slowly and cautiously, because there are many cases where T1 jobs write to MCDisk.

The plan, in principle, was to then proceed to move GroupDisk into DataDisk. However, it was unclear how groups would get what they should after the merge. Experience showed that without general rules and policies, the complications would multiply exponentially.

Discussion of how to address disk space requests

A trigger for this discussion was a request from SLAC physicists to get the entire JetMet NTUP dataset on disk at SLAC. The requestors had been advised to seek endorsement and possibly space from their physics group, but it raised the question of how to meet US-specific space needs.

Some talking points for the Disk Space discussion (circulated by Richard before the meeting):

  • The US has, and will have, substantial above-pledge disk resources (especially at T2s)
  • ATLAS is desperately short of disk space and is considering limiting access to datasets in ways that will make some physics analysis harder
  • Should the US make its beyond-pledge disk available to ATLAS as a whole?
  • Should we keep it separate for the benefit of US physics groups?
  • If we keep it separate, how do we decide who gets what (requests are already coming in)?
  • How much above pledge disk (and CPU) is available in the rest of ATLAS?

Torre commented on the budget situation: we should be very careful in commenting on or expecting beyond pledge resources.

LocalGroupDisk was discussed. This is likely to be an important or vital resource for US physicists. There were no space management policies for LocalGroupDisk (see Kaushik's caution above) and it was not clear what requirements or guidelines ATLAS placed on the provision of LocalGroupDisk. [After the meeting it was confirmed that ATLAS policy was that LocalGroupDisk should not be included as a pledged resource, and thus any provision of LocalGroupDisk required beyond-pledge funding.]

Kaushik volunteered Armen to report on data in GroupDisk and LocalGroupDisk within the US. It would also be valuable to find out about other clouds.


Richard noted that the CPU usage number (% of pledge) reported by WLCG seemed high in some cases. Kaushik also noted that SWT2 sometimes was credited with more CPU than it had on the floor. It was agreed that an investigation was necessary.

Action Items

  1. 1/28/2011: Armen: report on data in GroupDisk and LocalGroupDisk within the US. It would also be valuable to find out about other clouds.
  2. 9/24/2020: Kaushik: Organize the nature and timing of the effort to create an automated mechanism to give US-requested additional production priority access to US non-pledged resources. On hold - manual system works just fine for now

About This Site

Please note that this site is a content mirror of the BNL US ATLAS TWiki. To edit the content of this page, click the Edit this page button at the top of the page and log in with your US ATLAS computing account name and password.


pdf GROUPDISK_110204.pdf (9.8K) | ArmenVartapetian, 04 Mar 2011 - 14:51 |
pdf LOCALGROUPDISK_110204.pdf (9.3K) | ArmenVartapetian, 04 Mar 2011 - 14:51 |
Powered by TWiki
This site is powered by the TWiki collaboration platformCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback