Case Study: China Eastern Airlines achieves 100x performance improvements and 1.6B daily queries with MongoDB

MongoDB and Apache Spark at China Eastern Airlines: Delivering 100x Performance Improvements

China Eastern Airlines, one of the world’s largest carriers, needed to re-architect its fare calculation platform to handle dramatic online growth: selling 130,000 seats online translates to roughly 1.6 billion searches and fare requests per day (about 90,000 searches/sec), while their legacy Oracle system supported only ~200 searches/sec—necessitating roughly 100x greater throughput.

They built a new fare engine that pre-computes 180 million prices nightly using Apache Spark (applying business rules stored in MongoDB) and loads the results into MongoDB for the search app. The Databricks‑certified MongoDB Connector for Apache Spark provides efficient, locality‑aware data movement; on a cluster of fewer than 20 servers the platform supports 1.6 billion daily queries and delivers 15x higher performance with 10x lower latency versus the previous Oracle system, enabling the migration off Oracle and new application capabilities.

China Eastern Airlines

Chong Huang

Lead Architect