Using notrunning condition causes a loop

[L.LEONG]

I have a a jobs which cannot run concurrently. My problem is that once the success conditions are met, for some reason the jobs keep start over and over again.

 

For example, take these two jobs:

LRST-dist-reptmart-wrapup-clearing-fact01-load-etl-00d and LRST-dist-reptmart-setlmt-dep01-static-load-etl-00d

 

The start conditions for each job looks like this: (there are NO look backs in the condition statements)

 

LRST-dist-reptmart-setlmt-dep01-static-load-etl-00d

START CONDITIONS: (all "and")

Success from LRST-dist-dupes-passed-reject-load-jsb-00d

Not Running  LRST-dist-reptmart-mc-clearing-fact01-load-etl-00d

Not Running LRST-dist-reptmart-visa-clearing-fact01-load-etl-00d

Not Running LRST-dist-reptmart-setlmt-dep01-dyn-load-etl-00t

Not Running LRST-dist-reptmart-wrapup-clearing-fact01-load-etl-00d

 

 

LRST-dist-reptmart-wrapup-clearing-fact01-load-etl-00d

START CONDITIONS: (all "and")

Success from LRST-dist-deposit-requalify-jsb-00d

Not Running LRST-dist-reptmart-mc-clearing-fact01-load-etl-00d 

Not Running LRST-dist-reptmart-visa-clearing-fact01-load-etl-00d

Not Running LRST-dist-reptmart-setlmt-dep01-static-load-etl-00d

Not Running LRST-dist-reptmart-setlmt-dep01-dyn-load-etl-00t

 

However, once the success conditions are met, the jobs run like this:

 

LRST-dist-reptmart-setlmt-dep01-static-load-etl-00d ->starts

LRST-dist-reptmart-setlmt-dep01-static-load-etl-00d ->ends

check conditions on LRST-dist-reptmart-wrapup-clearing-fact01-load-etl-00d à conditions == true

 

conditions met -> run...

LRST-dist-reptmart-wrapup-clearing-fact01-load-etl-00d->starts

LRST-dist-reptmart-wrapup-clearing-fact01-load-etl-00d ->ends

check conditions on LRST-dist-reptmart-setlmt-dep01-static-load-etl-00d à conditions == true

 

conditions met -> run...

LRST-dist-reptmart-setlmt-dep01-static-load-etl-00d ->starts

LRST-dist-reptmart-setlmt-dep01-static-load-etl-00d ->ends

 

then the cycle repeats over and over.

      

What would you suggest to fix this problem?

[Chandru.Venkataraman]

You might want to add a look-back to the success() condition.

For example:

Job: LRST-dist-reptmart-wrapup-clearing-fact01-load-etl-00d

condition: S(LRST-dist-deposit-requalify-jsb-00d,0) & N(LRST-dist-reptmart-mc-clearing-fact01-load-etl-00d) ...

[L.LEONG]

Thank you. This solved my issue. A simple fix is always best.

[Paul Tayler]

Keep in mind that whenever the jobs in the condition statements change state to match the condition they trigger the dependent job to run. So when LRST-dist-reptmart-wrapup-clearing-fact01-load-etl-00d finishes it triggers LRST-dist-reptmart-setlmt-dep01-static-load-etl-00d and vice versa.  As Chandru suggests a look back should fix it.

[Paul Tayler]

If the look back on the success job does not suit you, try putting the LRST-dist-reptmart-setlmt-dep01-static-load-etl-00d and LRST-dist-reptmart-wrapup-clearing-fact01-load-etl-00d jobs into separate boxes (the dependency can still point to the jobs).  This will break the loop and these 2 jobs will not run without a starting event. eg start the box

[L.LEONG]

Thanks for the input!

[Daniel Shannon]

Another option is to use a virtual resource to serialize the jobs. You could create a renewable resource that has 1 unit. Each job would be defined to require 1 unit. When the jobs conditions are satisfied, one of the jobs will be started and hold the resource unit. The other job, if ready to start, would go into a RESWAIT state until the job consuming the resource completed.

[L.LEONG]

Thanks for your input. I'll need to keep this solution in mind.