Trying to get a series of jobs to repeat

[Bruce Jones]

              I have setup a series of jobs under a job parent,  that I need to have repeat throughout the day from 7 am until 9 pm.

Parent job  JOBA_D is setup to repeat every 30 minutes using the same occurrance, under the parent I have listed JOBA, JOBB, JOBC, JOBD each dependant upon one another with 'no repeat'.   JOBB also has a variable dependency,  because JOBA can get either a 0 exit code or 10.  If JOBA gets a 10 I don't want the rest of the jobs to continue.

When the jobs run if all get 0 exit code they all run fine,  and repeat every 30 minutes,  however if JOBA gets and exit code 10 then like I expect JOBB-D do not run,  however they also won't repeat every 30 minutes,  it looks to me like if the whole series does not finish then Tidal won't repeat the jobs.

We are on 5.3.1.260  just looking for suggestions on way to set this up.  The way I have the jobs setup now if how the Tidal manual says to do it.

Thanks    

[Marc Clasby]

I think you might be able to solve this if JOBA, JOBB, JOBC, JOBD are all under their own parents and the depdencies are set at the job parent levels. If it is all under the same group and you exist with 10 does the parent say waiting on children?

Marc

[Bruce Jones]

Marc,

You are right, if the first job get the exit code 10  then the parent waits on children.   I had the same idea as you, except what I did was split the first job out to its own parent, and the 3 other jobs to a seperate parent.  I will wait and see on tomorrows run if that fixes the problem,  of course I need to wait for the first job to get the exit code 10 again. 

Thanks for the response, sounds like we have the same idea.

[Bruce Jones]

Well after trying serveral things,  I found one way to make this work,  if you take Joba  and schedule it on its own and have it repeat the same instance and then take the other jobs and put into a group dependant on Joba then all the jobs will repeat, however depending on how long you group jobs run,  when the time resets for the next run will be calculated after the last job has finished in the group.   So I went back to what I know will work and that is to is run a new occurrance every 30 minutes, a little messy since you have a bunch of job instances sitting in the queue all day, but it works and all the jobs run at the time needed.