Unexpected behaviour with service tasks between parallel gateways

**Issue Description:**
The service tasks configured between parallel gateways are stuck abruptly without any exception.

**Example Use case:**

Here, both the parallel gateways are asynchronous & exclusive and service tasks A, B and C are also asynchronous and exclusive. However we observe a pattern where service tasks A and B are executed as expected and service task C is stuck without moving forward. ACT_RU_EXECUTION table is still pointing at service task C and not moving to suspension state. There are no timeouts from service task C and no trace to confirm that the service task has started it's execution.

**NOTE:** This issue is not consistent but happening quite a few times.

**Code:**
`

`