(aws-stepfunctions-tasks StepFunctionsStartExecution): (allow redrive instead of retry)

[kmolsen]

Describe the feature

Currently, we're able to add retries when we use a StepFunction to start another StepFunction. Redrive has recently been implemented for StepFunction executions, and it would be great if instead of retrying to whole StepFunction execution, we could redrive instead.

Use Case

It's frustrating when a StepFunction execution fails in the last task and may take a long time to rerun everything. This would eliminate that extra time and resources and allow the StepFunction execution to continue where it failed.

Proposed Solution

In the .retry method, add an option for redrive=True

Other Information

No response

Acknowledgements

• I may be able to implement this feature request
• This feature might incur a breaking change

CDK version used

v2

Environment details (OS name and version, etc.)

Ubuntu 20.04.5

[pahud]

I just read this blog post as well as the serverlessland emit-and-wait workflow sample but it's not clear if it is supported in cloudformation or if it could be defined in the workflow definition.

The redrive seems to be triggered programmically through AWS CLI, SDK or console. Do you have any idea how to define that in the workflow definition?

[abdelnn]

Redrive is a feature that is not defined in a state machine's workflow definition. It's therefore not represented in Cloudformation either. I don't believe that it belongs in the CDK at all, as it does not affect your infrastructure as code.

However, you should be able to specify a retry behaviour for each individual states. What you're looking for would be an enhancement/feature in the StepFunctions service itself