03_LambdaLayers

Lambda Layers

Custom Runtimes

• Through lambda layers once can support runtimes that are not supported by AWS.

• Allows languages that are not officially supported by AWS to be used to develop and deploy on Lambda.

• C++ rutimes is one such example. Some of the common ones can be found here.

Externalize Dependencies

• Dependencies which we use in our lambda code may not change often compare to our application/lambda code. So in order to avoid reuploading dependencies which seldom changes, one can leverage the lambda layers.

• As layers are externalized they can be reused across multiple function code.

Lambda Layers

• Additional libraries may be bundled and moved as a layer.

• Lambda can have upto 5 layers per function, with maximum deployment size limit of 50 MB/layer.

• One can create a lambda layer and publish it, if required lambda layer is not available.

• The packages added in layers are available in /opt directory during invocations.

• Native libraries should be first compiled in Amazon Linux.

• AWS SDK comes by default with every lambda function.

References

• Using Lambda Layers