MII Scheduler is getting stopped automatically

[former_member206397]

Hi All,

I have one MII scheduler which is posting data from MII Message buffer to SAP ECC through RFC call and the scheduler runes every 5 seconds interval. It is working as expected but the scheduler is getting stopped suddenly (let say once in a week or once in 2 or 3 months).  I am re-enabling the scheduler from scheduler editor and it is working again. My applications are running with MII 12.0 SP 12

Can anybody provide any suggestion.

Thanks & Regards

Chandan

[Former Member · Accepted solution]

Hi Chandan,

What the issue you have faced is quite similar to the issue we faced. So from that experience I can share few points,

1) It is not recommended for MII to set the interval of the scheduler in Sec. The recommendation is 2-5 min interval.

2) You can put a small logic in the BLS which is calling from the scheduler in the background. The logic is,

when scheduler call the BLS, within the BLS it will check the status of the BLS using transaction manager service, if it is in running state then stop schedule instance for the current cycle. and in the next cycle if the status is not running then complete the schedule instance.

by this logic, you can stop the back end queuing of instances in the MII internally which help you to reduce a huge load from the server (if this back end queuing run for minimum 30 min with the interval like 5 sec,then your MII server may crash based on the data size. At that time the status of the scheduler will be in running state continuously)

Hope it will help you,

Regards,

Suman

[former_member4529 · Accepted solution]

Hi,

I'll also suggest to use asynchronous transaction call inside the scheduled BLS to reduce the actual scheduled BLS execution time.

Thanks,

Dipankar

[former_member206397 · Accepted solution]

Hi Suman,

Thank you so much for your details reply. At first let me check with business if I can increase the Scheduler time interval from seconds to minute or I can restrict the number of message processing (from Message buffer) in each interval, so that the instance overlapping will not take place.

Thanks again

Chandan

[former_member206397 · Accepted solution]

Hi Dipankar,

Thank you for your suggestion. We are also in process of changing it to a  asynchronous transaction call.

Thanks again.

Chandan

[HariCS23]

Hi Chandan,

You can also try increasing 'ExecutorPoolMaxSize' and 'PercentageOfParallelismAllowed' values in Confgtool->managers->ApplicationThreadManager.

Thanks

Hari