This note focuses on design reasoning, failure behavior and operational evidence—the parts that matter in code review, system design and incident response.
Why this problem matters
A bill-management API must remain available during instance failure, deploy safely and preserve durable data. It also benefits from separating synchronous bill operations from asynchronous notifications so email delivery does not determine API latency.
A useful mental model
An Application Load Balancer distributes traffic across EC2 instances in an Auto Scaling Group. PostgreSQL RDS stores durable bill and user data. CloudFormation defines repeatable infrastructure. The request path remains synchronous through the database; notification work fans out through messaging and Lambda.
Design principles
The following principles are useful because each one creates a boundary that can be reviewed, tested and observed. They are not independent checkboxes: together they define the behavior of the system under normal load and partial failure.
Keep application instances stateless so Auto Scaling can replace them
Treat this as an architectural constraint rather than a cleanup item. Put the boundary in code, configuration or the data model so a reviewer can see exactly where it is enforced.
Use health checks that reflect readiness, not only process existence
The benefit becomes visible when timing changes under load or failure. Define the limit explicitly and make the fallback, rejection or recovery behavior observable.
Place RDS in private networking and restrict access by security group
Ownership matters here. The component that owns the invariant should also own the validation, compatibility rule and operational response when the assumption is violated.
Deploy immutable application bundles through a repeatable pipeline
Prefer the smallest mechanism that preserves correctness. Add sophistication only after measurements show that the simpler design cannot meet the workload.
Move email notifications to SNS/SQS/Lambda/SES so API success is decoupled from email delivery
Convert this principle into an automated test, deployment check or runbook step. Otherwise it will drift as dependencies, traffic and team ownership change.
How to validate: Verify the design with change sets, least-privilege review, deployment rollback and controlled infrastructure failure. A diagram is incomplete until recovery has been exercised.
Key trade-offs
Good engineering makes the cost of a choice visible. For this topic, the most important trade-offs are:
| Managed simplicity | Managed services reduce undifferentiated work but still require limits, IAM and failure planning. |
|---|---|
| Portability | More portability can introduce a larger platform surface and higher operating cost. |
| Blast radius | Stack and account boundaries should match how the system is deployed and recovered. |
Concrete example
The example below is intentionally small. Its purpose is to expose the control point or data flow that the design depends on, not to present a complete framework implementation.
Client → ALB → EC2 Auto Scaling → Express API → PostgreSQL RDS
└→ SNS → SQS → Lambda → SES
Infrastructure: CloudFormation
Delivery: CircleCI → S3 artifact → CodeDeployWhen applying this pattern, define what happens immediately before and after every durable boundary. That is where duplicate work, stale state, lock duration, timeout overlap or deployment risk usually enters the design.
Common failure modes
Failure modes are more useful than generic “best practices” because they describe the condition the design must survive. Review each one as a concrete test scenario.
- Storing sessions or uploaded state on one EC2 instance. The usual consequence is hidden backlog, duplicate work or state that can no longer be explained. Add a bounded guardrail and reproduce the condition under load.
- Using a health endpoint that succeeds before dependencies are ready. This often passes unit tests because the timing, cardinality or dependency behavior is too clean. Test it with realistic concurrency and an intentionally slow or failing dependency.
- Sending email synchronously inside the bill transaction. During restart or replay, the defect can turn a recoverable incident into inconsistent state. Preserve enough context to detect, stop and safely resume the workflow.
- Allowing deployment scripts to mutate infrastructure outside CloudFormation. The safest mitigation is to make the assumption explicit in a constraint, deadline, queue limit or state transition, then alert when the boundary is approached.
What to measure
Production behavior should be visible before a failure becomes a customer complaint. Metrics should connect a technical symptom to a workload, business state or recovery objective.
- ALB target health and 5xx rateUse this as an early saturation signal and define what healthy, warning and overloaded behavior look like.
- ASG desired versus healthy capacityBreak this down by service version, endpoint, partition or tenant so aggregate averages do not hide one failing path.
- RDS connections and storageCorrelate this with user-visible latency and error rate to distinguish harmless internal work from customer impact.
- Deployment success and rollbackTrack both the level and the age of the condition; an old small backlog can be more serious than a brief large spike.
- SQS age and Lambda failureReview this after deployments and failure drills so the dashboard proves recovery, not only steady-state health.
Interview-ready explanation
A strong explanation starts with the invariant: state what must remain true even when requests repeat, dependencies slow down or instances restart. Then describe the mechanism that preserves it, the failure mode that mechanism introduces and the signal that proves it is working.
For Designing a Highly Available AWS Bill Management Application, avoid listing tools first. Explain the workload and boundary, walk through the normal path, introduce one realistic failure and show how the system recovers. Finish with the metric or test that validates the claim. That structure demonstrates senior engineering judgment more clearly than naming patterns without context.
Review checklist
Use this checklist during design review, implementation planning or incident follow-up:
- Keep compute stateless.
- Use multi-AZ design where required.
- Define infrastructure as code.
- Separate notifications.
- Test instance and deployment failure.
A sound design is not the one with the most patterns. It is the one whose invariants, limits and recovery paths are explicit—and can be demonstrated.