What's new

This patch provides an updated BLAH to address some problems when SGE is used as batch system in the CREAM CE

Installation and configuration

• When updating, please use: yum update

  Please note that doing: yum update glite-CREAM you are not guaranteed to have all relevant RPMs updated.

• (Re)configuration is needed after the installation/update.
• If the new blparser is used (default option) it is just necessary to run yaim once (i.e. it is not necessary anymore to first configure the creamCE and then the BLparser). More information in: http://grid.pd.infn.it/cream/field.php?n=Main.CREAMAndBlparserConfiguration

Known issues

See: http://grid.pd.infn.it/cream/field.php?n=Main.KnownIssues

• First time that run BUpdaterSGE after a long time, if there are a lot of jobs in non final state (removed or completed) in the job registry, many hours can be required to synchronize BUpdaterSGE daemon with the actual status, because a qstat of every pending job is needed

Known issue identified during staged rollout:

The SGE configuration allows you to delete jobs in the batch system using SIGKILL or SIGTERM. This is something that it is up to the SGE Administrator to decide/configure, but AFAIK, SIGKILL is the default option. If you have configured SIGKILL to cancel a SGE job, BupdaterSGE recognizes the apropriate exit code, and logs the proper state to the Job Registry, and therefore, to CreamCE.

However, SIGTERM is not recognized because it produces an exit code for the job which is not properly recognized by BupdaterSGE. The job will be logged in the Job Registry as state 4 (completed) but with exit code 255, and is reported as ABORTED. SGE_utils developers are already aware of the problem.