A job execution can be repeated in any of the following exceptional situations (retries are not enabled by default):
- Errors are returned. In this case there are two options:
- All queries if any error is returned. The whole job is repeated when errors occurred during its execution. This option is only applicable to extraction jobs. In the case of ITP, JDBC, or VDP jobs, the job will only be retried if errors occurred in the extraction stage, and in the case of ARN, only if there are URLs with http access or Input/Output errors.
- Only queries which return an error. When there have been errors in the execution of the job (during the extraction stage), it only repeats those queries that have returned an error or have not started to be executed. This option has the same effects as executing the job from the “Scheduler” perspective using the “Start with state” option. For ARN jobs that do not execute multiple queries, this option has the same effect as the previous one (All queries if any error is returned). As in the previous option, it is not applicable to ARN-Index-type jobs.
- Exported documents are less than XX. The whole job is repeated when
- the number of exported documents is less than the specified threshold (option All queries if exported documents are less than: XX).
You can configure for how long retries are performed, in the event that after one or several retries the errors persist (you can also configure the Delay between executions, set in milliseconds and 0 by default). The options are as follows:
- Not limited. Retry while errors. Retries will be performed until no errors are returned.
- Limited to: XX. A maximum of XX retries will be performed even if errors persist.
When using retries and an exporter to a file (CSV or SQL) is configured, all the job retries are exported to the same file.