Print

Print


 
ATLAS SCCS Planning Oct 1 (and September), 2008
  -----------------------------------------------------------

  9am, SCCS Conf Rm A, to call in +1 510 665 5437, press 1, 3935#

 1. Tier-2 Production/Operation

    - Low activity during last month in ATLAS Grid production
      More Panda jobs expected with a new production release soon to be released. 
    - There were some issues with user jobs in pAthena having too 
       small cache size (10 MB) while accessing  XROOTD files. 
       The cache has been increased to 50 MB.  
 
 2. Local copy of Oracle CB and/or Proxy for Reconstruction jobs
 
    - SLAC has set up a solaris machine to host local copy of Oracle CB,
       but neither BNL or CERN is willing to provide distribution streams. 
       It will require a management decision. 
    - Secure Proxy downloading of conditions from Oracle DB at BNL is
      working fine with little latency.  However, it has become apparent that
      each batch job can not access the BNL CB, as it becomes saturated 
      at 100 jobs accessing simultaneously.  They will try to limit it to 
      50 at a time, but controlling GRID jobs to wait in queue does look 
      difficult.  If Tier-1 CB at BNL crashes, it could bring down Tier-0 
      as well.
 
 3. WT2 status and infrastructure
 
   - Current year CPU and disk spaces are fully installed and operational.
   - Plans for 2009 is not yet in the works.  With the role of SCCS changing
     it will be necessary to investigate price/performance of new systems
     purchase, and billing the power bills etc.  Some negotiation with US ATLAS
     is foreseen. A through study might require 2 FTE planning.  With the 
     technological cycles changes in mind, wait 6 months till decision.
  - Solaris 10 update to version 5 is still ongoing.
 
4. Network Monitoring machine request from US ATLAS
 
  - Wei will keep in touch with US ATLAS, as there are questions as to how 
     useful they will be, and also the security issues of  the OS system, 
     we will delay the purchase.
 
5. Interactive machine for SLAC-ATLAS users and 1/2 Thumper Disk space
 
  - Install ticket has been created by Wei for the 8 cpu Solaris machine 
    (RHEL4-64) and the Thumper to be shared 1/2 each with BABAR
  - Disk partitions will be in NFS. 
 
---------------------------------------------------------------------------