Awesome

Serverless test project

This project uses some central functionality of Serverless and shows how the alias plugin works.

You can deploy the available branches master and warmup-apig to different aliases. The branches have different configurations:

master

Simple Lambda function + DynamoDB table + DynamoDB event source

warmup-apig

The warmup branch simulates feature development in a project perspective as well as developer changes in the Serverless configuration.

The DynamoDB is removed in this branch, APIG endpoints are added and the warmup plugin has been added to serverless yml.

This branch can be deployed to a separate alias without destroying anything that is deployed on master. Branch development in master and warmup can even get more different over time.

sns-topic

This branch adds a SNS subscription event to one of the functions. It shows how the alias plugin will handle these when deployed to different aliases.

authorizer

This branch deploys a custom authorizer and attaches it to the func1 GET method. You can call GET https://XXXXXXXXX.execute-api.us-east-1.amazonaws.com/auth/func1 with an Authorization header set to an arbitrary value. The function will echo the incoming event object, that contains the results of the authorizer under requestContext.authorizer.

Deploying the system

Checkout the master branch and deploy the stage without specifying an alias (i.e. the branch gets deployed into the master/stage alias dev).

git checkout master
npm install
node ./node_modules/serverless/bin/serverless deploy

Checkout the warmup branch and deploy it to a different alias

git checkout warmup-apig
npm install
node ./node_modules/serverless/bin/serverless deploy --alias=warmup

git checkout sns-topic
npm install
node ./node_modules/serverless/bin/serverless deploy --alias=sns

Afterwards you will have everything deployed in your AWS account. The functions for the different aliases are labeled (aliased) with the alias name. See the documentation of the alias plugin for details.

Attention: The project will create resources in your AWS account. These will generate costs. You should delete the created resources (CF stacks) if you do not need them anymore.