Troubleshooting - Scheduler

Infinite loops

An infinite loop can be configured within a schedule where a dependent task is configured to return to the initial task, typically on exception. For example in the schedule below, tasks 2, 3, and 4 have been configured to start the Blue Prism process on an alternative resource when the previous task is an exception.

We could configure task 4 to call task 1 on exception. However, if a scenario emerges that causes the process to fail instantly, for example, missing credentials, then the number of failed process sessions and scheduler updates will cause the Blue Prism database to fill up with log entries. This puts other running processes at risk and could slow overall Blue Prism performance. The only way to remove such an infinite loop is to take down the Blue Prism application server. A better design would be for task 4 in our example to call a process that sends and email alert or create further tasks (5, 6, 7) to start the process.

I am unable to configure the scheduler in System Manager

Ensure that you have the correct permissions to perform this action. Contact your system administrator to check or assign the appropriate permissions. For more details, see User permissions.

I have made changes to the scheduler configuration in System Manager, but they don’t appear to be having any effect

Ensure that the Blue Prism server service has been restarted since you made your changes. The changes will not be activated until the service it restarted. Ensure that no sessions are running before restarting the service.

I am unable to create schedules and tasks in Control Room

Ensure that you have the correct permissions to perform this action. Contact your system administrator to check or assign the appropriate permissions. For more details, see User permissions.

Expected schedules have not executed

Ensure the following conditions are met:

  • The Blue Prism Server Service is running. This is required to trigger schedules.
  • The resource(s) on which the task(s) are scheduled to run were online at execution time. Ensure the resource(s) are visible and connected from the perspective of the application server.
  • The resource(s) on which the task(s) are scheduled to run are public resources.
  • The schedule is not configured to run on a date that doesn’t exist in the month (for example, configured to run on the 31st of the month and it was not executed in April).
  • The schedule is configured to run on the correct day of the week. Remember that in Blue Prism, the first day of the week is always Monday and the last day of the week is always Sunday.
  • The day on which the schedule has executed is configured to be a working day in the calendar being used by the schedule.
  • Check whether the resource on which the session has been scheduled has been subsequently moved to a resource pool. A schedule can only be run on a resource pool or a specified resource that is not part of a resource pool. A schedule originally configured to run on a specified resource which has subsequently been moved to a resource pool will fail. For more details, see Resource pools.
  • Multiple processes that are scheduled to run on the same resource concurrently are able to do so. If multiple processes are scheduled to run on the same resource and any of the processes use any objects with a run mode set to exclusive, the task will be terminated as the resource is too busy to run anything other than the exclusive process.
  • The domain on which the target resource is hosted is included in the application server’s domain suffix search list. The Blue Prism scheduler runs on the application server. When a schedule is triggered, the scheduler sends a TCP/IP message to the resource to start the schedule using the host name of the resource. If the resource is hosted on a different domain to the application server, the host cannot be found if the domain on which the target resource is hosted on is not included in the application server’s domain suffix search list.
  • When configuring schedules it is necessary to consider that the execution is based on the time zone for which the application server is configured. Additionally, when there are multiple Blue Prism servers deployed for a given environment that are configured to execute schedules, it is important to specify a common time zone. The time zone of the runtime resources and the interactive clients is not taken into account as part of the scheduled execution.
  • The event logs from the application servers have been reviewed.

I am not able to run a schedule manually when in a different time zone

When manually running a schedule from Control Room via the Run Now option, its execution may fail if your local time zone (which is also the time zone of the Blue Prism interactive client) is different from the timezone of the Blue Prism application server.

This can be remedied by manually changing the timezone in your Windows settings to match the application server or by configuring a terminal server or jump box and running the interactive client from that jump box in the same region as the application server.

In Citrix VDEs, a Citrix administrator will need to set the time zone for the interactive client machines to the same timezone as the application server. Once this has been configured correctly, you will then be able to run the schedule manually via the Run Now option.

A number of sessions started via a schedule have terminated unexpectedly

Check whether the sessions all belong to the same task. A failure of any session within a task will cause the whole task to fail. Therefore, if multiple sessions are scheduled, a failure of any of these sessions will cause all the other sessions to terminate. Deselect the Fail fast on any errors option.

Queue items are automatically marked as an exception for no obvious reason

As described above, the Fail fast on any errors setting is designed to fail all sessions in a task if any one of them fails. When this happens, any queue item currently being worked will be marked as an exception when the session terminates. When investigating this type of queue item exception, the logs may suggest the session failed for no apparent reason and without handling the queue item correctly. The reason for this is the scheduler terminated the session abruptly because the Fail fast on any errors option was enabled.

A process started by the scheduler is not performing as expected

Ensure the start-up parameters have been input correctly in the task. For more details, see Start parameters .