ccj jun2000

Opening of PHENIX Computing Center in Japan (PHENIX CC-J): 

Opening of PHENIX Computing Center in Japan (PHENIX CC-J) Yasushi WATANABE RIKEN / RBRC Jun 9, 2000 PHENIX Core week

The PHENIX Computing Center in Japan (PHENIX CC-J) is now open to whole PHENIX Collaboration for Simulation projects.We accept proposals for simulation officially from now: 

The PHENIX Computing Center in Japan (PHENIX CC-J) is now open to whole PHENIX Collaboration for Simulation projects. We accept proposals for simulation officially from now

Contents: 

Contents What is the CC-J? Current status How to use Summary

What is the CC-J?- http://ccjsun.riken.go.jp/ccj/ -: 

What is the CC-J? - http://ccjsun.riken.go.jp/ccj/ - Regional computer center for scientists in Japan and Asian countries Serve for RHIC physics activities for scientists in Japan and Asian countries Major analysis center for Spin physics A Major Simulation Center for PHENIX Open for whole PHENIX collaboration Major CPU resource for PHENIX No CPU resource for simulation at BNL-RCF 70% of CC-J CPU resource is assigned for Simulation

An important concept of the CC-J: 

An important concept of the CC-J Identical (as much as possible) with RCF from user point of view Login/Data server: SUN E450, Solaris Farm: PC clone (AltaCluster), RedHat6.1 HPSS: pftp, data carousel (not yet), rftp (in consider) Batch queue: LSF We think much of durability of LSF for such large facility AFS: Using mirrored NFS service as AFS for the farm Users can work as RCF once apply phenix_setup.csh

Current status: 

Current status Login Server: SUN E450 (ccjsun) NFS Server: SUN E450 w/ 288GB + 3.2 TB Disk Farm: AltaCluster 6 boxes (48 nodes = 96 CPUs) MHz*CPU: 450*32, 600*32, 700*32 70% (for Simulation) = 85 CPUs of 450MHz equivalent PHENIX@RCF: 138 CPUs (CRS: 92 CPUs, CAS: 46 CPUs) LINUX: RedHat6.1, Kernel 2.2.14 + NFSv3 patch HPSS: 100 TB (tape capacity) COS: Simulation, Raw, DST, Analysis, Backup

Slide7: 

CC-J Room SUN E450 * 2 3.2 TB Disk (RAID5) 48 nodes (96 CPUs) of PC clone: AltaCluster Gigabit Ethernet: between HPSS and SUNs

Slide8: 

HPSS Room IBM RS6000-H50 * 5 w/ ~1TB cache disk STK Silo w/ Tape (50GB) * 2000 = 100 TB RedWood tape drive * 4

CC-J Organization: 

CC-J Organization Planning and Coordination Office (PCO) Responsible for operation of the CC-J Manager: T. Ichihara Technical manager: Y. Watanabe Scientific program coordinator: H. Hamagaki, H. En’yo Computer scientist: N. Hayashi, S. Sawada, S. Yokkaichi CC-J liaison at BNL: N. Saito PHENIX Simulation Coordinator: Charlie Maguire Optimization of simulation projects

How to use  - http://ccjsun.riken.go.jp/ccj/doc/usersguide/ccjusersguide.html - : 

How to use - http://ccjsun.riken.go.jp/ccj/doc/usersguide/ccjusersguide.html - If you have Simulation project Make a proposal Talk with PWG: PWG has a responsibility of the project Define responsible staff(s): YOU might be a person Bring the proposal to Simulation Coordinator and CC-J Planning and Coordination Office (PCO) If no conflict or optimization is exist, you will be assigned resources (CPU#…) as the proposal. Execute job Responsible person for job management (You?) is asked to stay at Wako. Report the outcome of the project to PCO

Application form of Proposals: 

Application form of Proposals http://ccjsun.riken.go.jp/ccj/forms/app.html

Proposal of the computing project: 

Proposal of the computing project Required information Title & the name of Physics Working Group Purpose and physics merit of computing Description of computing items Required resources CPU time, DISK storage, HPSS storage Responsible staffs for job submission, for QA, and the liaison of the project Single person can act for the three roles Target date for the completion

Responsibility of PWG: 

Responsibility of PWG PWG has responsibility of the computing projects Submit proposals accepted by own group Quality Assurance of software Job execution Report the outcome to PCO PWGs have to assign responsible staffs for these jobs Liaison for the project Responsible person for QA Responsible person for Job management Single person can act for the three jobs

Slide14: 

Schedule 1998 1999 2000 2001 2002 RBRC (BNL) R&D for CC-J RIKEN Wako Phase 1 Phase 2 Phase 3 1/3 scale 2/3 scale Full scale Prototype of CPU farms Data Duplication facility April April April April CC-J review at BNL (Dec. 1998) HPSS Software/Hardware Installation (March 1999) (Supplementary Budget) CC-J starts operation at 1/3 scale (April. 2000) Full scale CC-J (Mar. 2002) CC-J construction CC-J front-end at BNL April CC-J Working Group formed (Oct. 1998) PHENIX Exp. at RHIC

Summary: 

Summary We (CC-J PCO) strongly encourage all PHENIX collaborators to use CC-J for Simulation projects. The CC-J may not be well suited yet. However using and responding make the facility improve. Remember, the facility power will be reach more than 2 times powerful until next March, and could be reach 4 times more until March/2002. Suggestions and comments are very welcome phenix-ccj-pco@rarfaxp.riken.go.jp