3

I'm well aware of the lambda function deployment package size limit is 50 MB(in case of compressed .zip/.jar) with a direct upload and 250 MB limit (uncompressed) via upload from S3.

What I'm not clear is how lambda deploys the package from S3?

  1. Is it on the each invocation of the lambda function?
  2. Will there be any cost associated data transfer between S3 to lambda function?
Schparky
  • 429
  • 4
  • 10
Rex
  • 521
  • 3
  • 8

1 Answers1

3

  1. No, after you upload the deployment package, it's saved in the function and layer storage of your AWS lambda account which has a default limit of 75GB. On each invocation of the lambda function, the deployment package will be pulled from there.

  2. Since the deployment package is not pulled from S3, it will not incur any data transfer cost.

jellycsc
  • 10,904
  • 2
  • 15
  • 32
  • Do you, by chance, have a reference for this? I have seen things that allude to the fact that Lambda doesn't use the S3 bucket after deployment, but nothing seems conclusive. Thanks for the answer regardless. – Schparky Apr 03 '23 at 02:01
  • @Schparky Unfortunately no official docs have mentioned this. The best way to get a conclusive answer is to test it yourself. The easiest way to prove this is to delete the bucket after the lambda function deployment and see if the lambda still works. If it still works, then no cost is incurred on deleted S3 resources. – jellycsc Apr 03 '23 at 04:17