- General information
- Data readiness
- Grid and Cloud
- Infrastructure
- Machine Learning
- Offline Software
- Production
- S&C internal group meetings
- Test tree
Running Embedding
Updated on Fri, 2007-02-16 21:10. Originally created by starembed on 2007-02-15 18:01.
Under:
This page describes how to run embedding jobs once the daq files and tags files are in place (see other page about embedding production setup).
Jobs are usually run by user starofl but personal accounts with group starprod membership will work, too (but test first as the group starprod write permissions typically are not in place by default).
The script to submit a set of jobs is submit.[user].pl. The script should be modified to submit an embedding set from the the configuration file
Basics:
Embedding code is located in production specific directories: ~starofl/embedding/P0xxx. The basic job submission template is typically called submit.starofl.pl in that directory.Jobs are usually run by user starofl but personal accounts with group starprod membership will work, too (but test first as the group starprod write permissions typically are not in place by default).
The script to submit a set of jobs is submit.[user].pl. The script should be modified to submit an embedding set from the the configuration file
~starofl/embedding/[Production]/setup/[Particle]_[set]_[ID].setup
where
[Particle] is the particle type submitted (Piminus for GEANTID=9, as set inside file)
[set] is the file set submitted (more on this later)
[ID] is the embedding request numberTest procedure:
The best way to test a particular job configuration is to run a single job in "DISK" mode (by selecting a specific daq file in your submission). In this mode all of the intermediate files, scripts, logs, etc., are saved on disk. The location will be under the "data" link in the working directory. You can then go and figure out which script failed, hack as necessary and and try to make things work...Details:
»
- Printer-friendly version
- Login or register to post comments