Speaker
Dr
Sanjay Padhi
(University of Wisconsin-Madison)
Description
With the evolution of various Grid Technologies along with foreseen
first LHC collision this year, a homogeneous and interoperable Production
system for ATLAS is a necessity. We present the CRONUS, which a Condor
Glide-in based ATLAS Production Executor. The Condor glide-in daemons
traverse to the Worker nodes, submitted via Condor-G or gLite RB. Once
activated, they preserve the Master-Worker relationships, with the worker
pulling the production jobs sequentially until the expiry of their
lifetimes. The initial startup glide-ins not only ensures a guaranteed
ATLAS software environment but also provides a homogeneous large pools of
resources across different grid flavors. The structure of CRONUS and how
it handles job management, resource selection, security etc. is described.
The requirement of a secondary worker from the same Glide-in daemon for
DATA Transfer or any other maintenance jobs is also discussed.
| Submitted on behalf of Collaboration (ex, BaBar, ATLAS) | ATLAS |
|---|
Primary author
Dr
Sanjay Padhi
(University of Wisconsin-Madison)
Co-author
Dr
Rodney Walker
(TRIUMF)
