AWS Cost Design and Operations Policy

This policy outlines the procedures and configurations for managing AWS costs, resource hygiene, and business continuity for "My Roadmap".

1. AWS Budget Alerts

To prevent unexpected charges, budget alerts are configured via AWS CDK in the backend.

Alert Name	Threshold (USD)	Type	Status
Monthly Budget	$5.00	Actual (100%)	Automated (CDK)
Forecasted Alert	$5.00	Forecasted (100%)	Automated (CDK)

Configuration (Automated)

The alerts are defined in web/amplify/backend.ts using aws-cdk-lib/aws-budgets.

Setting the Notification Email

The notification email is retrieved from the environment variable BUDGET_NOTIFICATION_EMAIL.

How to set it:

1. Log in to the AWS Amplify Console.
2. Select your app -> App settings -> Environment variables.
3. Click Manage variables and add:
   • Variable: BUDGET_NOTIFICATION_EMAIL
   • Value: your-email@example.com
4. Re-deploy the application for the changes to take effect.

2. Cost Monitoring Strategy

• Monthly Review: On the first Monday of every month, review the AWS Cost Explorer monthly report.
• Anomaly Detection: Check for unusual spikes in AppSync query volume or DynamoDB read/write capacity.
• Tagging: Ensure all project resources are tagged with Project: MyRoadmap for accurate cost allocation (handled by Amplify Gen 2 default tagging).

3. Resource Management & Hygiene

CloudWatch Logs

• Retention Policy: All log groups created by Amplify/Lambda must have a retention period of 14 days.
• Reasoning: Standard logs are only needed for immediate troubleshooting; long-term storage adds unnecessary costs.

Database Operations

• DynamoDB Backups: Enable Point-in-Time Recovery (PITR) for the production Task table.
• Cleanup: Regularly review and delete unused S3 buckets or Amplify playground branches.

4. Disaster Recovery (DR) & Backup

Component	Strategy	Recovery Point Objective (RPO)
Frontend	Code-based (GitHub Actions Redeploy)	< 1 hour
Backend API	IaC (Amplify Gen 2 / CDK)	< 1 hour
Database	DynamoDB PITR (Continuous Backup)	1 second

Recovery Procedure

In case of regional failure or accidental deletion:

1. Re-deploy the backend using npx ampx pipeline-deploy (or via GitHub Actions).
2. Restore the DynamoDB table from the latest Point-in-Time backup in the AWS Console.
3. Update the amplify_outputs.json in the frontend if endpoint URLs change.