Case Study: Kuaidi (Greater China’s largest car‑service app) achieves real‑time driver–passenger matching and massive scale with MongoDB

Leaf in the Wild Scaling China’s Largest Car Service App with MongoDB

Kuaidi is Greater China’s largest car‑service app (founded 2012) with about 100 million users, 5 million drivers and up to 8 million ride requests per day across 300+ cities. The company needed low‑latency, real‑time location tracking and an order archive that could scale beyond the limits they hit with MySQL after exceeding 100 million records.

Kuaidi migrated the location service and order archive to MongoDB, using geospatial indexes and sharding across four geographic regions (7‑node replica sets per region) to keep data close to users and ensure high availability. The deployment now serves ~50,000 ops/sec (80:20 reads:writes) and nearly half a billion documents, delivering the horizontal scale, performance and operational simplicity Kuaidi required.

Kuaidi

Ouyang Kang

Chief Architect