An APS application task fails in Operations Automation (OA) with this error message:
Another operation on instance of application / service is already in progress.
The task Initiating operation on instance with id
<ID> or Finalizing operation on instance with id
<ID> that previously was scheduled for the problem APS application instance was canceled or removed.
When OA performs an operation on an APS application, the following three tasks are created with the same mutex (in the same queue):
Initiating operation on instance - Marks an APS application instance as being (re)configured in the OA database
Executing configuration scripts - Performs actual operations on the APS application instance
- Finalizing operation on instance - Marks an APS application instance in the OA database as being completely (re)configured
Therefore, if the task Initiating operation on instance or Finalizing operation on instance is canceled, the application instance will be marked in the OA database as still being (re)configured and successive operations under the application instance will fail. This is because the previous operation still is running according to the OA database.
Please never cancel or delete the Initiating operation
<ID> or Finalizing operation
<ID> APS tasks in OA. This will lead to inconsistent data in the OA database and to the problem described in this article.
Canceled OA tasks can be re-executed, so find all the tasks related to the problem APS application instance in Task Manager in the OA Provider Control Panel and re-run them. Depending on the version of OA you are using, you can find these tasks in:
- OA 5.4: System Director > Task Manager > Canceled Tasks
- OA >= 5.5: Operations > Tasks > Canceled Tasks
Tasks may be found by searching for the same Subscription ID or queue name as the failed task, or by the name of the APS application through the "Name" column. In the latter case, please use asterisks around the name, like:
You can find additional information in this Knowledgebase article: How to find OA tasks for particular application instance.
Open a ticket with the Odin Support Team if the problem is not resolved after all the canceled tasks are re-run and executed successfully.