The Compute Farm and Batch System at SLAC
Last Update: 03 Apr at 13:57:02 PDT 2006
The Compute Farm and Batch System Documentation
High Performance gives a detailed explanation of the Compute Farm and LSF batch system used at SLAC.
lsfintro and lsbatch describe the components that make up the batch system at SLAC. These components works together with the Compute Farm to schedule and process batch jobs. A postscript copy of the user guide lsf6.1_using.pdf is located in /afs/slac/package/lsf/currdoc/pdf/manuals
The Staging Facility
Stagein is a method where tape data is first copied to a staging disk where it can be read by one or more users. Copying the data to the staging disk is fairly fast and frees the
tape drive for other requests. It also saves wear and tear on tapes, especially if the data is to be read many times. After the user has selected certain information from the stagein data, he or she may want to write the selected information out to tape
using Stageout and Stagewrt. Stageout allows users access to large amounts ( >500MB) of temporary disk
space and Stagewrt facilitates ease of writing data out to tape or disk. Stageclr is used to clean up the staging disks when the user has finished. The section Example of Using
Staging has some examples. The staging facility is a component of the Compute Farm system.
Example of Using Staging
transferring data using Staging.
Maintained by Charlotte Hee. Send suggestions, comments, or additions to
chee@slac.stanford.edu, 415-926-3353
|