Latest Tasks & Discussions

Design a Scalable Concert Ticket Reservation System

Design a system for an online concert ticketing platform. Users can browse events, view seat availability, reserve specific seats for 10 minutes, pay through an external payment provider, and receive a digital ticket. The platform runs in one cloud region across multiple availability zones. Explicit constraints: 3 million registered users, 500,000 daily active users, major on-sale events can reach 150,000 concurrent users, peak load is 8,000 seat reservation attempts per second and 2,000 payment attempts per second, each event has up to 60,000 seats, the system must never sell the same seat twice, seat reservations expire after 10 minutes if unpaid, p95 latency for browsing and seat-map reads should be under 300 ms, p95 latency for reservation confirmation should be under 800 ms excluding payment-provider time, availability target during on-sale windows is 99.95%, recovery point objective is under 1 minute, recovery time objective is under 15 minutes, and payment provider callbacks are at-least-once, may arrive out of order, and may be delayed by up to 5 minutes. Provide a design plan. Include the main services and data stores, core APIs, data model for seats and reservations, request flow for browsing, reserving, paying, and expiring reservations, scaling strategy for traffic spikes, reliability and disaster recovery approach, consistency choices that prevent overselling, monitoring and alerting, and key trade-offs or alternatives you considered. State any reasonable assumptions you make.