Error Management

When a job encounters an error or is cancelled, it can behave in two ways:

  • Stalled (default) — The job remains active in the scheduler but pauses due to the error.
  • Stopped — The job fully stops execution.

This behavior depends on the the ria-boot.properties com.ria.converter.stopJobOnError property, or the stopJobOnError Task parameter:

  • false — the job becomes stalled, which is the default and recommended setting
  • true — the job will be stopped.

Stalled jobs

By default, a failed task causes its job to enter a stalled state and appear under the Stalled Chain group in the Jobs Activity page. Although the run status may show the job as ended with an error, it is still considered running by the Oracle DBMS Scheduler and can be verified through the ALL_SCHEDULER_RUNNING_CHAINS view. To avoid unnecessary resource usage, the issue should be resolved and the job restarted, or the job should be stopped, with the option to restart it later.

Error Detailes for a stalled job

To determine the cause of failure, follow these steps:

  1. Open the Active Run Detail page by selecting the stalled job.
  2. Review the chain steps and their statuses.
  3. Click the error number to open the Log Message Detail dialog and view the detailed error message. This message usually matches the run status details and is sufficient for troubleshooting.
  4. Optional. If more information is needed, check the Run Status page.

This feature is only available for stalled jobs; if the job is stopped, the error message may be generic and require full checking.

Restarting a stalled job

Once the error has been identified and corrected, the job can be restarted from the following pages:

  • Jobs Activity — click the resume icon. This will restart all the failed steps if the job executed steps in parallel and more than one had failed.
  • Active Run Detail — click the resume icon. This will only restart the failed step. If you need to restart several steps, you have to restart them individually.
  • Log Message Detail — click the Restart Step button. This will only restart the failed step. If you need to restart several steps, you have to restart them individually.

Stopping a stalled job

To stop a stalled job, click the stop icon. This will end the run in the DBMS Scheduler and free up resources.

Once a job has been stopped, click anywhere on the Jobs Activity row for the stopped job, excluding the orange highlighted job name and any icons, to open the Run Log Detail page. This will show various statistical details, as well as the error message that caused the failure.

To restart the job from this stopped state, simply navigate back to Jobs Activity and click the run icon.

Stopped jobs

Jobs enter this state when the legacy stop option is enabled by setting either the com.ria.converter.stopJobOnError property or the stopJobOnError Task parameter to true.

In this mode, any failed Task causes the entire job to be stopped, and in the Jobs Activity page the job is shown in a stopped state. Unlike stalled jobs, the error message that caused the failure is typically only available in the Run Status and must be reviewed there to determine the root cause.

To restart the job, click the play icon. The job will skip successfully completed steps and resume from the step that failed.

A job can also be forced to restart from the beginning by selecting the option for the job’s last run in the Run Status page. When the job completes successfully, its status is updated to Complete.

Scheduled job error handling

If a scheduled job encounters an error, handling depends on the com.ria.converter.stopJobOnError property or the stopJobOnError Task parameter.

If both the property and parameter are false , the job enters a stalled state. It can then be restarted after the issue has been identified and resolved. Once the job completes successfully, it resumes its regular schedule. If a stalled job is manually stopped using the Stop Job button, it will also resume its regular schedule. This setting is recommended for scheduled job executions.

If either the property or parameter is set to true, the tool forcefully stops the job, and it returns to its regular schedule. In this case, the Scheduler UI does not clearly indicate that the job failed, so error detection relies on email notifications or manual monitoring of the Run Status in the main UI.