Overview
Festival spikes caused app/API slowdowns and order failures. The platform implemented layered caching, autoscaling, and queue‑based workflows to stabilize.
Key Changes
- CDN + edge caching for menus/assets; server‑side caching for hot categories.
- Database optimization: read replicas, slow query fixes, and connection pooling.
- Message queues for order events (placed → accepted → picked → delivered); resilient retries.
- Dispatch optimization: geofencing and batched assignments to riders.
Outcomes
- P95 latency down 45%; failed orders reduced by 70% during peak hours.
- More stable rider utilization; fewer customer support tickets.
Lessons (Unit 3 lens)
- Caching + queues are fundamental under bursty, location‑dependent traffic.
- Separate read vs. write paths; monitor capacity in advance of seasonal peaks.
Chapters covered
- Web performance and scalability (3.2–3.4)
- Mobile considerations for high‑traffic events (3.5)

