Handling High Traffic Scenario at Amazon

How would you handle a scenario of large traffic causing the website to crash? (caching, scaling, load balancing)

Asked in the round 2 LPs+scripting+troubleshooting stage. Troubleshooting/system-design concepts discussion.

Explain how you would diagnose the failure, stabilize the service, and prevent recurrence. Address caching, horizontal scaling, load balancing, overload protection, and graceful degradation. If the website includes an ML-backed endpoint, explain how model serving and feature dependencies affect the design.

1. State the first operational checks and containment steps.
2. Propose a resilient request path and scaling strategy.
3. Define caching, load-balancing, and rate-limiting behavior.
4. Explain ML inference fallbacks if relevant.
5. Describe monitoring, testing, and capacity planning.

Constraints: preserve data correctness, avoid cascading failures, control cost, and maintain security while traffic is elevated.