Resource level retry logic

[corymhall]

Hello!

• Vote on this issue by adding a 👍 reaction
• If you want to implement this feature, comment to let us know (we'll work with you on design, scheduling, etc.)

Issue details

Each AWS service has different API limits and require different retry logic. The CCAPI schema does not contain this information so we currently have global retry logic which applies to all resources.

For example the `` resource can have a large propagation delay when creating. The terraform-aws-provider has it configured to retry for up to 5 minutes. hashicorp/terraform-provider-aws@7a8669e/internal/service/ec2/vpc_route_table_association.go?plain=1#L80. It doesn't make sense to make 5 minutes the global retry limit though.

Affected area/feature

[corymhall]

Another example pulumi/pulumi-cdk#277 where ApiGateway can only delete 1 api per 30 seconds.

[flostadler]

Another example is SQS queues: #1983

You need to wait 1 minute after deleting a queue before creating one with the same name again