MongoDB Atlas Setup Guide

Overview

MongoDB Atlas is a managed MongoDB service. It is an alternative to DynamoDB for projects requiring a document database with rich query capabilities, aggregation pipelines, or when a project needs to remain cloud-agnostic.

Atlas Project Structure

Atlas uses a hierarchy: Organization → Project → Cluster → Database → Collection.

Level	Example
Project	myapp sandbox, myapp production
Cluster	Main-NDB-Project
Database	ndb_project
Collection	users, orders, etc.

Cluster Types

Type	Best For	Scaling
Serverless	Variable traffic, no capacity planning	Auto-scales, pay per operation
Dedicated	Predictable traffic, specific performance needs	Manual or auto-scaling
Shared (Free/M0)	Development and prototyping	Fixed resources

Serverless is recommended for most Superloom projects.

Configuration

• Provider: AWS (for co-location with other services)
• Region: Same as your primary AWS region
• Continuous Backup: Enabled (for production data protection)

Connection

• Authentication: Username/password
• Network Access: Allow access from anywhere (or restrict to known IPs)
• Connection String Format: mongodb+srv://user:password@cluster-host/?retryWrites=true&w=majority

Local Development

The framework's Docker Compose provides a local MongoDB instance:

docker compose -f docs/dev/docker-compose.yml up mongodb -d

Default local connection: mongodb://localhost:27017/app_dev

Security

• Create separate projects for sandbox and production
• Use dedicated database users (not the admin account) for application access
• Enable audit logging for production clusters
• Store all credentials in SSM Parameter Store, actual values in __dev__/secrets/