Hephy Innsbruck


Infrastructure


1) Activities

  • operating a Top-Level BDII for Central Europe ROC
  • ROC test site and competence centre for CREAM-CE
  • storage of reprocessed ATLAS Cosmics data in GridKa Cloud: 10% of AODs/DPDs in Innsbruck (about 500GB - 1TB)‏
  • pilot testbed for ATLAS DDM

2) Hardware

  • Computing Elements
    • lcgCE with Torque/Maui Batch System and SL 4.8
    • glite-creamCE with Torque/Maui Batch System and SL 4.8
    • 28 WN: 2 x Quad Intel(R) Xeon(R) CPU L5420, 2.5 GHz, 2 GB RAM
    • 9 WN: 2 x Dual Intel(R) Xeon(R), 3.0 GHz, 2 GB RAM

  • Storage Elements
    • SUMO Raid with 48 hard drives of 1 TB size each
    • Starline Easy raid with 16 hard drives of 1 TB size each

  • additional Grid Servers
    • gLite Monitoring System Collector Server (grid-mon)
    • Berkeley Database Information Index (BDII) (grid-bdii)
    • gLite Workload Management System (grid-rb)
    • T2 Squid (atlassq)

3) Planned upgrades

  • increase of 40 – 80 TB, depending on available hard drive size
  • 80 additional cores (WN) for ATLAS
  • 1 GBit/s connection to ACOnet



ICT Network Layout


ICT network


Tier-2 Layout


Tier-2 Layout



Hephy Vienna


Infrastructure


1) Computing Element

  • standard LCG CE with Torque/Maui Batch System and SLC 4.7
  • present about 400 CPUs will be increased up to 1000 CPUs in summer 2009
  • new Worker Nodes (WN): SUN blades, Intel 2 x Quad processors

2) Storage Element

  • DPM on SLC 4.7 with XFS file system, planned to move to dCache under Solaris with ZFS file system
  • Supermicro raid with 45 hard drives of 1 TB size each
  • present HW RAID-6, will be upgraded to a software raid with ZFS file system
  • present about 100 TB, upgrade up to 500 TB until summer 2009


CMS PhEDEx - Cumulative Transfer Volume


PhEDEx