Architecture

Project Description

When stepping into Serverless 🌐, finding the right architecture 🏛️ can make all the difference.

I recently built a Serverless API on AWS 🚀implementing the Services Design Pattern. In the Services Pattern, a single Lambda function can handle a few jobs that are usually related to a single entity of the data model (e.g., User 👤).

All CRUD operations on the User data model are performed on the single HTTP endpoint using different HTTP methods.

For this to work, you can have a 'small' router (HTTP method) at the beginning of your Lambda code.⚙️

It's different from the Microservices Pattern where each job is isolated within a separate Lambda function.💡

Benefits of the Services Design Pattern: ✔️

• Fewer Lambda functions to manage ✔️
• Reduced cold starts ✔️
• Team autonomy 👥
• Faster deployments 🚀

Drawbacks of Services Pattern: ❌

• More complicated debugging ❌
• Requires a router ❌
• Bigger function size ❌

Tech Stack I utilised: 💻

• Go
• Terraform

AWS Services I utilised: 🛠️

• AWS API Gateway
• AWS Lambda
• AWS DynamoDB
• AWS CloudWatch