If you work with Oracle Identity Manager, you should already know that process tasks can be re-tried based on the “Retry Period in Minutes” and “Retry Count” parameters which are both set in the task definition screens.
The Open Provisioning (OP) tasks are triggered when the Task Timed Retry schedule task is run. But the Task Timed Retry task does not trigger OP tasks when they pass the retry count threshold.
Imagine the following scenario:
You have the Retry Count set at five (5) for Process Task (“Change Job Title”) within a connector. The task fails (maybe due to bad password, lost connectivity, or similar). The task will retry based on the schedule setting for a maximum of five times. If the underlying problem (e.g. loss of connectivity) is not corrected within the specified time limit, the count limit will be reached and the OP task will ultimately fail to run successfully and remain in an open state.
There is a system maximum on the number of retries which may be reached relatively quickly (depending on your retry interval). For example, if the interval is set to retry every 10 minutes, you may hit the maximum retry count in about one week.
The FishEye Group Solution:
The FishEye Group has developed a solution that captures the OP tasks which have hit their max retry limit and remain open. The solution leverages a custom task scheduler to identify those tasks and automate the retry process. You may also choose to initiate another separate process such as an email notification or a network connectivity test to raise alerts. This solution is available to system integrators as well as to Oracle OIM customers. Let us know if you'd like details on how to implement this solution in your OIM environment.
