Helge Meinhard (CERN-IT)
10/27/09, 1:15 PM
This talk will present an update on R&D activities around storage at CERN. The main focus will be various activities around iSCSI technology, but an update will also be presented on the Lustre evaluation project.
Pierre Emmanuel Brinette (CNRS-CCIN2P3)
10/27/09, 1:40 PM
IN2P3 Computing Center has been using HPSS as a Mass Storage System since 1999. There has been no major system upgrade since 2005 and IN2P3 still runs HPSS 5.1. This version is no more supported by IBM and doesn't include the T10K-B drive support. In June 2009, the system has been upgraded to HPSS 18.104.22.168. This upgrade implies major changes (DCE removal, DB2 and systems upgrade). This...
Jason Hick (LBNL)
10/27/09, 2:05 PM
The High Performance Storage System (HPSS) has served the DOE community for high performance archival storage for the past fifteen years. It specifically serves the HEP community at LBNL/NERSC by providing archival storage from the PDSF system. This presentation will provide a brief overview of how HPSS works, what its current more unique features are, and what our plans our for our next...
Jonathan Schaeffer (CCIN2P3)
10/27/09, 2:30 PM
TReqS is our Tape Request Scheduler. Based on BNLBatch, its goal is to get between dCache and HPSS and to reorder the files requests. Since May of this year, a first implementation hit our production system. We will present here : - the problematic of tape access for LHC experiments - the solution we implemented - TReqS in its production environement, our 5 month experiment with the beast
Peter van der Reest (DESY)
10/27/09, 2:55 PM
At DESY we have installed a testbed for Grid and storage related issues. One of the first trials we did was running a ALICE Proof job against data held in a distributed NFS 4.1 service based on dCache. Also trails were made against an industrial system, of which I am not allowed to speak yet. Comparisons we made with regard to scaling in PROOF and IO performance in comparison to NFS3 To make...