Hey all, one highly requested feature has been to support some way of specifying retry options for resource operations: https://github.com/pulumi/pulumi/issues/7932 We now have an RFC up for a proposed way to support this: https://github.com/pulumi/pulumi/discussions/20950 We'd love to hear any comments, concerns or thoughts to ensure this will cover everyone's use cases before we commit to the plan.

This is very exciting. Our team has been looking for a mechanism like this for a while. We'll take a gander. Thanks for all the work you guys do ❤️

Would this include built-in retry mechanisms for providers? (the

azure-native

provider has been notoriously bad due to the API returning retryable errors constantly…) I would hate to have to manually add retries to every resource.

Would this include built-in retry mechanisms for providers?

No that's a per-provider problem, but we'll be looking into ways to make this easy to apply to whole stacks so you can deal with all azure resources at once.

> Would this include built-in retry mechanisms for providers?

No that's a per-provider problem, but we'll be looking into ways to make this easy to apply to whole stacks so you can deal with all azure resources at once.

That's unfortunate. I think the

azure-native

provider could benefit quite a bit from built-in retries (retries that I don't have to manually inject in any way).

I'm not on the azure team but my limited understanding is they consider that to have about as many drawbacks as positives. It is something that keeps coming up though, I'll have a talk with them about it again.

😄

@thousands-memory-85842 look and post in #C01HEMG55T2. On #CB36DSVSA is the wrong place for that inquiry.

This will be useful for some provider APIs with rate limits: https://github.com/pulumi/pulumi-aws/issues/5939