When a command job is defined without the std_out_file clause in the job definition, the output of command will be logged in the agent spool directory.
For example in:
This may cause the agent file system to get full and causing agent to crash when the volume of the job output exceeds the disk capacity.
If the following parameter is already set in "agentparm.txt" file : agent.spool.success.autocleanup=true ,
the spool cleanup will be activated only after the job is completed successfully.
While the disk space issue may occur while the job is still running. On the other hand, client may want to remove the standard output logging to the spool only for some jobs.
In fact, the System Agent will keep the standard output under the spool directory until the job is completed.
If you have some jobs for which you do not want to get the standard output logged into the spool directory,
one possible solution is to set the std_out_file to /dev/null in your job definition.
This will allow to avoid getting the data logged in the spool directory