What you also may want to try is putting the process within "batch mode" We just ran into a problem where firming and scheduling po's did the same thing. But if we put it in batch mode we got an error message that a record was deadlocked after a while. This was because when running it in client mode default ax had a retry statement inside of its try so until the records became unlocked it would try to update it instead of erroring out like it should. Once we had the deadlock error msg I just searched LCS and found out there was an KB hotfix available for both R2/R3 for this exact error.
This scenario only helps if you don't have any custom code that is being ran in your process. But it may be another option to try out.