Hi folks! We are trying to demonstrate Pulumi in the context of deploying lambda functions. We potentially have "a lot" of lambdas to deploy (500+) and each have multiple resources. We are using s3 backend for the prototype to have something that is production-grade. We are getting taxed with a 2 seconds delay at least per resources using the S3 backend which lead to very long deployment times. Anybody have tricks to make the s3 backend faster with large stacks (1000+ resources) ?

Hello @little-fall-55550, we have an experimental flag which might help in the round trip time save all the intermediate state files. It does come with its own drawbacks, so test this thoroughly on non-production stacks first. It’s experimental, so you won’t find it in the docs on our website for now. https://github.com/pulumi/pulumi/blob/a7a806e815decf08701c4bd2fbbc39b4699a2371/CHANGELOG.md?plain=1#L2058-L2068

Hi @limited-rainbow-51650 ! Thanks for the pointer, this is exactly what I was looking for. I was planning as a workaround to manually bring the stack locally for application / destruction so this is way simpler.