Testing the Production Software Releases - 15.6.9.Y
WARNING: This page is obsolete.
As a check on the quality of the production software releases installed at the Tier-2s, the fullchain CSC monte carlo has been run on specific releases, and the output from the MC has been compared to output from other Canadian sites. The recipe for this check is to use glite job submissions to produce 50 t tbar events with the transforms
Evgen_trf.py,
csc_atlasG4_trf.py,
csc_digi_trf.py, and
Reco_trf.py.
The glite commands used are (eg.):
glite-wms-job-submit --vo atlas -a -r ce1.triumf.ca:2119/jobmanager-lcgpbs-atlas csc_fullchain_allsites.jdl
glite-wms-job-output --dir /tmp <job-id>
(See the attached scripts for details on what is actually run). Of course, for valid comparisons, identical scripts are run on each Computing Element for a given release.
| Site |
Execution node |
GCC version |
| TRIUMF CE1 |
Intel(R) Xeon(R) CPU 5160 @ 3.00GHz |
4.3.2 |
| TRIUMF CE2 |
Intel(R) Xeon(R) CPU E5540 @ 2.53GHz |
4.3.2 |
| CA-ALBERTA |
Intel(R) Xeon(R) CPU L5420 @ 2.50GHz |
4.3.2 |
| SFU - Snowpatch |
Intel(R) Xeon(R) CPU X5355 @ 2.66GHz |
4.3.2 |
| SFU - Bugaboo |
Intel(R) Xeon(R) CPU E5430 @ 2.66GHz |
4.3.2 |
| Toronto - SciNet |
Intel(R) XEON(TM) CPU 2.60GHz |
4.3.2 |
| Victoria |
Intel(R) Xeon(TM) CPU 2.80GHz |
4.3.2 |
| Australia |
|
4.3.2 |
============================================================
15.6.9
--
JosephSteele - 30 Apr 2010