How to auto restart the job/step for a particular abend?

[Kishore_mf]

Hi,

I want my job step to restart automatically, if it fails for one particular abend.

For eg., Job step to restart of its own when it abends of U0777 user abend and NOT for any other abends.

Please help.

[Robert Sample]

If your site is using a job scheduler such as Zeke or CA-7 or Tivoli, work with the site support group to implement your requirement.

If your site is not using a job scheduler, what you want to do is not possible -- "restart automatically" implies some form of job scheduler.

[dick scherrer]

Hello and welcome to the forum,

If the job just abended, why would running it again succeed

Hopefully, there is no thought of letting it abend a few thousand times until somehow it actually runs . . .

[Akatsukami]

Hopefully, there is no thought of letting it abend a few thousand times until somehow it actually runs . . .

If it prevents one software engineer from getting called out at 0-dark-thirty...

[Robert Sample]

Hopefully, there is no thought of letting it abend a few thousand times until somehow it actually runs . . .

WAY back in my undergraduate days, when the primary printers for the entire school were a couple of dot matrix printers, an engineering student tied up both printers for an hour printing repeated compiles of a FORTRAN program. When asked why he ran the same program, with the same syntax error, through the compiler so many times, the reply was "I hoped that maybe one of them would work".

[shub2204]

Hopefully, there is no thought of letting it abend a few thousand times until somehow it actually runs . . .

If it prevents one software engineer from getting called out at 0-dark-thirty...

@dick sherrer
There are many jobs which are just sending data from mainfraime to other systems. If the ping to receiving server fails at a instant then job will abend. After few seconds if the server is Up, if it can restart itself it will not abend again.

[Robert Sample]

@dick sherrer
There are many jobs which are just sending data from mainfraime to other systems. If the ping to receiving server fails at a instant then job will abend. After few seconds if the server is Up, if it can restart itself it will not abend again.

Mr. Scherrer passed away about 5 years ago. I doubt you'll get a response from him on your comment.

Furthermore, if there is a network outage (among other possibilities), the restart will continue to fail until the network problem is corrected -- which could be minutes to weeks depending upon the situation. So the concern remains about how many times the restart is allowed.