We are letting CloudFormation to auto generating resource names for

Function

and

Bucket

because we want them to be unique. • For

Function

, if we hardcode the function name, the CloudWatch log group name will also be hardcoded. And because CloudWatch log groups are retained after you remove them, so if you remove the stack and redeploy, the deployment will fail with the error

log group with the same id already exist.

• For

Bucket

, the name needs to be globally unique. So it’s safer to let CloudFormation to auto generate them.

And when CloudFormation auto generates a name, it often has the stack name in them.

Let me know if that makes sense.

That said, the auto generated

Function

names are pretty ugly, and we are thinking of fixing it - https://github.com/serverless-stack/serverless-stack/discussions/236 I’d love to have you chime in the thread and share your ideas.