Data connection errors are critical errors in RPE/RRDG and by default any such failure will cause the document generation to abort immediately. There are however scenarios where this default behavior is not desirable:
- data containing known broken links
- peak load times when requests might sporadically be rejected by the server
For scenario #1 the solution is to set the ignore data source configuration errors property to true in the metadata section of the document specification.
Update: ignoring data sources can result in incomplete hence incorrect documents. If you are using this option make sure you review what was ignored to decide if the ignored data source was a broken link or actually some other error.
Even if ignored, failed data source requests are important errors and should be reviewed by the document generator. Since the RPE logs can be quite large and searching for errors in the log text can be difficult, RPE logs all failed data source requests in the problems view as well. This way you can audit the document generation and decide if the document can be used.
- data request retry count – the number of times to retry a failed request. Default value is 0
- data request retry interval – the time ( in milliseconds) to wait between 2 consecutive retries. Ignored if data request retry count is 0
When it comes to RRDG, to the best of my knowledge these properties are not currently used or accessible in any of the existing RRDG implementations.