Start (submit) a mvs job from within a Unix shell script

Rafael Longo · Posted: Tue Sep 04, 2007 12:05 am

Hello, everyone,

Is there a way to start (submit) a mvs job from within a Unix shell script?

dick scherrer · Posted: Tue Sep 04, 2007 5:37 am

Hello,

Suggest you post much more complete details.

From the way you've posted the question, "maybe" comes to mind as an answer.

Rafael Longo · Posted: Tue Sep 04, 2007 8:44 am

Hi, Dick. Ok. Let's see:

I have a process, which has a portion that runs on a unix box.

At the end of it, the unix ksh script transfers the file to mainframe, creating a dataset. This is where the mainframe portion of the process should begin.

I would like to start a job on mainframe right after transferring the file. Can I do it from the same unix ksh script?

Let me know if this is enough...

Thanks,
Rafael

superk · Posted: Tue Sep 04, 2007 1:45 pm

Some possible scenarios come to mind:

1. Make a call to an object in your job scheduling system from the shell script, telling it to schedule a job on the mainframe host system.

2. Have a mainframe scheduling system schedule that uses a dataset trigger. One you upload the file and meet the requirements for the dataset trigger, the job is scheduled.

3. Use a file-watch product on the mainframe host which can trigger a scheduler activity when a particular activity is recorded.

4. Use FTP to either "put" JCL from the Unix system to run on the mainframe host, or to "get" JCL from a host-based library and run on the mainframe host. See this link.

Rafael Longo · Posted: Tue Sep 04, 2007 6:59 pm

Hi, Superk,

Thanks for the suggestions.

I like the second alternative. Looks like it fits exactly what I need... Triggering the job schedule when the file is uploaded.

Could you, please, provide any guidance on how to set up a dataset trigger? Any code sample, example or documentation would do...

Thanks!
Rafael

enrico-sorichetti · Posted: Tue Sep 04, 2007 7:43 pm