Advanced
Stop the cascade
At 10:00 the recommendation service will go down for an hour. Last time, the whole store went down with it. Keep the store selling while recommendations are unavailable.
The storefront calls the recommendation service on every page. Java threads block while they wait, and immediate retries hammer the failing service.
What your design must achieve
- Availability ≥ 99.5 %
- Latency p99 < 1500 ms
- Infrastructure cost per month < 4,000 $ bonus