Serve static work at the edge
Let CloudFront answer cacheable requests close to the visitor while the private S3 origin remains inaccessible from the public internet.
An interactive architecture brief for static edge delivery, on-demand compute, durable asynchronous work, and controlled failure paths.
Select a scenario to see which parts of the system participate and what the operating model must account for.
Rendering architecture diagram…
Each service earns its place by changing delivery, isolation, durability, or operating responsibility.
Let CloudFront answer cacheable requests close to the visitor while the private S3 origin remains inaccessible from the public internet.
Use event-driven functions for dynamic behavior instead of maintaining an application server that waits for traffic.
Keep functions replaceable. Durable state belongs in services designed for persistence, concurrency, backup, and recovery.
Bound retries, isolate exhausted work, expose useful operating signals, and give callers a controlled result when dependencies fail.
The architecture avoids a permanently running application server. It still consumes managed services whose pricing, quotas, and abuse exposure need review against the actual workload.
Estimate with named traffic assumptions, then verify against current AWS pricing before deployment.
| Area | Cost behavior | Boundary |
|---|---|---|
| DNS | Hosted-zone baseline plus queries | Route 53 is not scale-to-zero. |
| Static delivery | Storage, requests, and transfer | CloudFront and S3 usage remains billable. |
| Dynamic path | Requests, duration, and capacity consumed | API Gateway, Lambda, DynamoDB, and SQS charge by their own units. |
| Operations | Logs, metrics, alarms, and optional controls | Retention and services such as WAF must be chosen deliberately. |
The public reference repository treats deployment code, operating checks, and rollback notes as one deliverable.
Lint, type-check, test constructs, and synthesize the infrastructure template.
Review the generated change set, IAM changes, replacement risk, and estimated service footprint.
Assume a scoped CI role through OIDC and apply the reviewed infrastructure change.
Run smoke checks, inspect alarms and logs, then record the deployed version and rollback point.