How to Create a Batch Job

The Unix Operating system allows applications to be shared by workstations, mainframes and PCs. Unix commands are short abbreviations of longer words. Programmers can use Unix-SAS to create batch jobs (multiple processes that can run with the execution of a script) in order to run multiple programs after logging out. Programmers typically create a file that contains the commands they wish to run at a later point in time. Batch jobs can be terminated individually or in their entirety at any point while they are running.

Instructions

1. Type in the command "at -f sasfile 5 am" at the command line and hit enter. This lets the system know that a file called "sasfile" will be run at 5 a.m. Change the time specification according to your needs.

2. Type in "atq" on the command line and hit enter to find out what jobs are currently in the queue waiting to move forward. Each job has a process ID next to the job name.

3. Type in "atrm pid" to terminate an individual job and hit enter. "pid" is the process number beside the job name in the listing created in Step 2. Note: The "remove batch job" steps are important If you set off a batch and then discover you have made an error and have to cancel/re-do the whole process. This is a key part of running a successful batch job.

4. Type in "atrm userid" at the command line and hit enter to terminate the entire batch of jobs. "userid" is the userid you use to log into the system.

References

• NYU - STERN: Unix Basics

Resources

• University of Texas: Unix and Linux Command Reference Guide