IBM Mainframe Forum

I have a job, job 1 that creates dataset a.b.c(+1) and it submits job 2 which uses a.b.c(00). The problem i have is that job 1 can run several time in one day and if the system is running slow the jobs could possible back up. So say job 1 runs twice in a row and job2 now runs twice but both or waiting to run. Once they run both jobs will use the same dataset (00). I don't have any kind of scheduler or CA7 so i can put a dependence on the jobs. Help any suggestions?

Without a job scheduler, there is not really a way to guarantee job execution order -- that is the purpose of a scheduler. If you have JES3, you can use DJC to force job 1 to run before job 2. But if a second job 1 is submitted before job 2 runs and clears the DJC network, the second job 1 will be placed on operator hold and must be manually released.

Your best bet is probably to create a file with the last GDG number processed and have a program as part of job 1 read that file and submit job 2 to the internal reader with an absolute generation number instead of using the relative generation. If there is a requirement that the generations be processed in order, you either use a scheduler or you report back that the assignment is not possible as given.

Hello,

How long do these jobs run? What causes job1 to be submitted? What do the 2 jobs do?

Is there any reason that job2 could not process any/all available data rather than only 1 file?

Have you tried using internal-reader to submit the second job? just a thought...