Question

A development team maintains 15 Lambda functions written in Python that all depend on the same set of third-party libraries (about 40 MB total) plus a shared internal utility module. Currently each function's deployment package bundles these dependencies, making packages large and forcing every function to be redeployed whenever a library is patched. The team wants to reduce deployment package size and manage the shared dependencies in one place. What is the most appropriate approach?

Accepted Answer

Publish the shared libraries and utility module as a Lambda layer and attach that layer to each of the 15 functions.

Answer

Move the shared libraries into an Amazon S3 bucket and have each function download them into /tmp at cold start.

Answer

Store the dependencies in Lambda environment variables and load them at runtime in each function.

Answer

Increase the memory allocation on each function so the larger deployment packages load faster.