Support runtime provided.al2 for debugging

[L4B0MB4]

When developing a go lambda including a external extension one has to switch to provided.al2. If you want to debug the lambda with

sam local start-api --host 0.0.0.0 -d 5858 --debugger-path dlv --debug-args "-delveAPI=2"

it only works until it builds the docker image.

How are we not supposed to debug this ?

Also (as a side note) I could have sworn it worked at one point. But after multiple colleagues had this problem I found myself unable to reproduce a working debugger as well - so it might be my dementia.

Any plans on supporting this in the (near) future ?

[CoshUS]

Instead of using the go emulation image, provided runtime will pull the provided lambda emulation image which does not support debugging. We currently don't have plans on supporting this. Tagging it as a feature request.

[alijarjis-pol]

I'd also like to see this feature implemented. Its very annoying that there is no way to debug a Go Lambda locally at the moment.

[shearn89]

Developing on a mac and generally working on ARM lambdas, not being able to debug is certainly a problem!

[ravitejaavv]

Any update on this issue

[ShankaranarayananBR]

Any update on this issue??

[mcblair]

We use lambdas behind API Gateway, and use gin proxy. In order for us to debug our SAM functions in al2 runtime, we've implemented an abstraction that determines if running in AWS - if so, we call lambda.Start with gin proxy, if not, we call a localStart function that starts gin directly.

I'm not aware of any alternatives, but this at least allows us to debug a single function.

[sbilello]

@mcblair could you share a sample repository with an hello world lambda that are you able to debug in local in GoLand or VScode in an M1 Mac please?