We're currently evaluating this issue for a fix in a future patch or service pack.
We've identified two causes of this issue that can be addressed without a code-fix:
- The Queue or business process in question is actually nested within another queue that is also running that process. (i.e., the disabling of the schedule actually is working correctly, and the queue was being invoked as a step within another queue) Evaluate whether or not this queue should be running while nested as a step of the other queue, and handle appropriately.
- A SQL server failover or database migration can result in two copies of the SQL Server Agent job both running. Disabling the old SQL Server Agent will prevent further running in that scenario.