[Lambda] Rust SDK for Durable Functions

[BDQ]

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Tell us about your request

Rust SDK support for Durable Functions - even basic support for just step and wait would be great, and waitForCallback would be a stretch goal!

Which service(s) is this request for?

Lambda Durable Functions

Tell us about the problem you're trying to solve. What are you trying to do, and why is it hard?

We use Rust exclusively in our Lambda runtimes, and would like to take advantage of Durable Functions soon.

Are you currently working around this issue?

We're not, we're currently blocked from adopting Durable Functions in our Rust codebases.

[jeastham1993]

I raised an issue for this on the Rust runtime crate as well if you want to upvote it there 😄 I think that's where the actual implementation would need to happen.

aws/aws-lambda-rust-runtime#1066

[lmammino]

I raised an issue for this on the Rust runtime crate as well if you want to upvote it there 😄 I think that's where the actual implementation would need to happen.

aws/aws-lambda-rust-runtime#1066

I was actually trying to understand this today and I believe the runtime API doesn't have any new function related to durable functions which makes me believe the work happens at the lambda service control plan. My current understanding is that once a checkpoint is created the control plane is responsible for shutting down (or somehow stop) the current instance that is executing the event.

This makes me think that once there's support for a Rust native durable functions SDK, the current Lambda runtime for Rust should be just enough.

Can anyone at AWS confirm this and/or provide more documentation? I am genuinely curious to understand how the stop/restart functionality is implemented under the hood (since i think it might have implications on other Lambda topics such as multi-concurrency with Manged Instances or even if we always incur in a cold-start on every retry).

Thx :)

[alessandrobologna]

@lmammino I have just posted a comment in aws/aws-lambda-rust-runtime/issues/1066 as an example that actually proves your point, that the rust lambda runtime is enough to implement a durable sdk. It's purely experimental of course and certainly unsupported.

Note: it's hacky right now to deploy it, since al2023 is not a supported runtime, so you have to deploy it as a "nodejs24.x" runtime (I guess python3.14 would work too) and then set AWS_LAMBDA_EXEC_WRAPPER to /var/task/bootstrap. See the ugly hack here: https://github.com/alessandrobologna/lambda-durable-execution-rust/blob/main/examples/template.yaml#L20-L35