Automatic Retry (x times) in case of error

[vivien57]

Hello everyone,

As many of us have done with SSIS and SQL Server jobs, I would like to know if it is possible, in the event of a Data Pipeline error during a scheduled run, to automatically restart it some time later?

I sometimes have flows that are in error because of a connectivity error or something else, but when we restart them in the morning, the execution is successful.

So I'd like to be able to have it restart automatically 1 or 2 times at night so that the feed isn't in error in the morning.

Thanks in advance for your feedback,

Have a nice day,

Vivien

[nilendraFabric]

Hello @vivien57 

did you tried retry policies?

Fabric allows you to set retry policies at the activity level within a pipeline. This is useful for handling temporary failures such as network glitches or service unavailability. You can configure:
• Retry Count: The number of retry attempts (e.g., 3 retries).
• Retry Interval: The delay between retry attempts (e.g., 30 seconds).

[vivien57]

Hello @nilendraFabric 

Thank you for your feedback.

It has to be set on each component (and not all components have a setting), so I'm looking for a method that allows it to be set for the entire flow (potentially at schedule level).

[Anonymous · Accepted solution]

Hi @vivien57,

Fabric’s built-in retry policy is limited to individual activities, and not all activities support it. As of now, there’s no native global retry at the pipeline or schedule level in Microsoft Fabric Data Pipelines.

Regards,

Vinay