lambda-layer-example

This is a sample SAM template for creating a Lambda Layer using Python. For an example of utilizing this layer in a lambda, see https://github.com/mcbanderson/lambda-utilize-layer-example

The following is an overview of what is provided in this repository.

.
├── README.md                   <-- This README file
├── .travis.yml                 <-- Example travisci pipeline
├── src                         <-- Source code for our lambda layer
│   ├── __init__.py
│   ├── my_module.py            <-- The module your layer will provide
│   ├── requirements.txt        <-- Module requirements
├── template.yaml               <-- SAM Template
└── tests                       <-- Unit tests
    └── unit
        ├── __init__.py
        └── test_handler.py

Requirements

• AWS CLI already configured
• SAM CLI installed
• Python 3 installed

Setup

Creating Your Layer

The first step is to create the module or package you want your Lambda Layer to provide. In this case we have created a module, my_module.py, which contains a single function, my_function(), which simply returns the value of the numpy constant numpy.e.

Once you have created your module/package, we need to specify an AWS SAM LayerVersion in our SAM template (template.yaml).

Resources:
    MyLayer:
        Type: AWS::Serverless::LayerVersion
        Properties:
            CompatibleRuntimes:
                - python3.6
                - python3.7
            ContentUri: dist/
            Description: A sample lambda layer
            LayerName: my_layer

Notice that we have specified ContentUri: dist/. For this example we will copy our module to dist/ before we package it, but you can use whatever directory name you like, as it has no effect once the Layer has been packaged.

Packaging and Deploying

Before we can package and deploy our Layer, we need an S3 bucket where we can upload our packaged Lambda Layer - If you don't have an S3 bucket to store code artifacts in, then this is a good time to create one:

aws s3 mb s3://<BUCKET_NAME>

Next we need to create a directory to put our module in for packaging. The name of this directory doesn't matter, but for this example we'll name it dist.

mkdir dist

We also need to create a directory within dist/, named python, in which we will place our module and its dependencies. THE NAME OF THIS DIRECTORY MUST BE python, otherwise our module will not be accessible to our Lambda. See the Lambda Layer docs for more info.

mkdir dist/python

Now we will copy our module to dist/python/, as well as any pip packages required by our module.

cp src/my_module.py dist/python
pip install -r src/requirements.txt -t dist/python

Now we are finally ready to package and deploy our Layer. To package, run:

sam package \
    --template-file template.yaml \
    --output-template-file packaged.yaml \
    --s3-bucket <BUCKET_NAME>

And to deploy it, run:

sam deploy \
    --template-file packaged.yaml \
    --stack-name <STACK_NAME> \

And that's it! Once your Stack is deployed, your Layer will be ready to use.

Using Your Layer

You may have noticed that we provided an SSM Parameter in our template.yaml. This is included so we can easily include our Layer in our Lambdas. As mentioned earlier, you can go to https://github.com/mcbanderson/lambda-utilize-layer-example to see an example of how to utilize your new Layer.

Cleanup

In order to delete our Layer, we just need to delete the CloudFormation stack that was created. To do so use the following AWS CLI command:

aws cloudformation delete-stack --stack-name <STACK_NAME>