Validate every change
The deployment model starts by compiling code from Composer or Git repositories. Every change can be validated safely in Beta with automated API, UI, and Performance tests before it reaches Production.
Platform overview
Architecture and CI/CD on AWS for Magento
Learn how the deployment pipeline and AWS architecture work together to deliver an isolated, scalable, and observable Magento platform.
Pipeline
Automate with control
After the required Beta tests pass, promotion can continue automatically or wait for manual approval. Backward-compatible Production changes use a blue-green strategy for zero-downtime releases and rollback.
Environment isolation
Beta and Production run in separate AWS accounts. Teams can experiment more freely in Beta without exposing Production data or access, while Production can retain stricter access policies.
The Pipeline view brings together deployment status, test history, logs, CloudWatch dashboard links, change and build identifiers, and rollback controls. Planned capabilities include Feature Toggles, custom Lambda functions, Production tests, and canary releases.
Architecture on AWS for Magento
Edge
CloudFront and WAF
HTTPS traffic enters through Amazon CloudFront for caching, compression, TLS termination, and AWS WAF protection against DDoS and common web attacks. The remaining infrastructure stays in private subnets.
Application
ALB and ECS Fargate
An Application Load Balancer distributes requests across Magento containers on ECS Fargate. Auto scaling adjusts capacity with load, while CodeDeploy orchestrates blue-green deployments and safe rollbacks.
Shared files
Amazon EFS
Shared directories such as Magento media are stored on Amazon EFS and mounted into every container, keeping files consistent as tasks are replaced or scaled horizontally.
Managed Data Services
Amazon RDS
MySQL storage with automated backups and plan-dependent point-in-time recovery.
Amazon OpenSearch
Catalog search and indexing for the storefront for Magento.
Amazon ElastiCache
Redis-backed application caching and session workloads.
Amazon MQ
RabbitMQ-compatible asynchronous message processing.
Amazon SES
Managed delivery for transactional application email.
High availability
Services are configured to remain operational when an individual node fails.
Observability and Alerts
Infrastructure and application services continuously publish logs and metrics to Amazon CloudWatch, including NGINX and PHP-FPM output. A main dashboard aggregates key service metrics, and custom dashboards or metric filters can be added when required.
- Default CloudWatch alarms include deployment rollback signals.
- Custom alarms can track service-level indicators and objectives.
- AWS X-Ray can provide end-to-end request latency and tracing.
Alert notifications can be delivered by email, SMS, or other configured channels so the appropriate team can respond when a deployment or service requires attention.