Have a look at IFS Cloud - Scheduling demystified (23R2) | IFS Community, chapter Scheduling Optimization Dataset. The database tasks should be created automatically.

Thanks @Alexander Heinze 

I actually have that document open.

However I am getting slightly confused which task should be used for scheduling!

We need to use a database chain task, as we need a custom task to execute first before the core PSO Load/Change database tasks execute.

The custom task decides what should be included/excluded for scheduling and updates the work task.

If I go to Scheduling Optimization Datasets and click View schedule, it takes me to database task “Send PSO Integration dataset updates” which I assume is the correct one to use for the PSO changes? (Screenshot in the document differs which references database task “Send Scheduling Optimization Dataset updates” - This does not exist in 23R2!)

There is another scheduled task “Schedule Plan Changes” that uses method Svcsch_Task_Allocation_API.Schedule_Plan_Changes that is executing every 2 minutes, and looks to be using arguments DATASET_ID, MIN_DIFF and WORK_ASSIGNMENT_STATUS. 

This may be related to UPDATE_PLAN_CHANGES parameter in Scheduling Configuration Parameters.

What is this doing and how does this differ to the above?

I just checked my 23.2.1 instance and “View Schedule” takes me to exactly the DB Task Schedule mentioned in the document:

That’s the task that sends regular updates to PSO (in my case every 10 minutes). Cannot shed light on any of the database tasks I’m afraid...