When suspening a process instance which has running jobs, these jobs will always run to completion, but then fail with an OptimisticLockingException.
Is there any way to avoid this? I would like the lob to simply complete regularly, and the process instance to continue with the next job after it is resumed. We have many activities with non-transactional side-effects, where retries are undesireable as long as the activity itself finished without errors.