We're using the KCRT_REQUEST_UTIL.move_request_workflow in a PL/SQL Stored Procedure to move multiple requests from one step to another. This is all working as desired, with one exception...
The process makes it through around 715 records successfully (process_phase = 5 and process_status = 7 on kwfl_transactions_int). After the 715th or so record, the process_phase = 1 and the process_status = 1.
The procedure is designed to be rerunnable, so the 2nd time through the records that were not processed the 1st time through (the 1 and 1's) get picked up and move successfully (5 and 7) until around the 715th record, then 1 and 1's again.
I have to rerun the SP several times to get through all of the records, 715 or so at a time.
Does anyone have any thoughts on what's going on here? I should note, nothing get's inserted into the knta_interface_errors table - the records do not error, they just don't complete.
Are you doing these all with the same group_id then calling the interface to run them all as a batch, or doing them one-by-one? I know there is some dynamic SQL in those interface procedures so you may be running up against the limit of how many it can do at one time.
In the version we have (PPM 7.5) it doesn't look like sending them all as a single batch is even an option - group_id is not one of the parameters passed to the procedure. It assigns a new one per each instance of it being executed.
The way we're doing it is executing the move_request_workflow for each request that needs to be moved. We may have a couple thousand to move, so it will be called a couple thousand times.