#DECLUTTER APP CODE CODE#

#DECLUTTER APP CODE CODE#

Why? Most of the code changes are going to happen in the repo with the Lambda functions. It often ends up easier to keep them in separate repos. So if you are creating a Serverless API backend, you'll have roughly two types of resources: your infrastructure resources and the Lambda functions for your API endpoints. They need to be able to share code and config easily between themselves. While, resources like DynamoDB, Cognito, or S3 are less likely to do so.Īdditionally, your Lambda functions have all your business logic code. For example, your Lambda functions and API Gateway endpoints get updated fairly frequently. But often you can draw a line between the resources that get updated frequently and the ones that aren't. This usually makes it so that your AWS resources and business logic code end up being closely coupled. And make our deployment process as simple as possibleįor starters, Serverless effectively requires you to adopt the infrastructure as code paradigm.Make it easy to create and configure new environments from the repos.Share code and config easily between services.But we want to pick a setup that allows us to: We are not going to be focusing on their specific details here. There are some very passionate arguments on the advantages and disadvantages of a monorepo vs multi-repo setup.

Where a service is configured using a single serverless.yml file. In this post we'll share some of the best practices for organizing and managing large Serverless applications.Ī quick reminder on the definitions before we get started. As your Serverless app starts to grow, you reach a point where you are trying to figure out how best to organize it.

Declutter app code