Case Study: BIGO achieves stable, scalable real-time messaging with StreamNative and Apache Pulsar

BIGO Embarking on the Road to a Stable and Scalable Real-time Messaging Experience with Apache Pulsar

BIGO, a gaming and social media company, needed a stable, scalable real-time messaging platform to support large-scale ETL, analytics, and recommendation workloads, while reducing the maintenance burden and performance limits they had experienced with Kafka. It worked with StreamNative to evaluate and optimize Apache Pulsar, especially for its pub/sub model and integration with Flink.

With StreamNative’s help, BIGO put Apache Pulsar into production and built a real-time messaging architecture for thousands of topics with independent schemas. The result was strong: Pulsar now processes tens of billions of messages per day at 2 to 3 GB per second, has cut hardware costs by nearly 50%, and has supported a stable ETL flow for nearly two years across more than 10,000 topics.

BIGO

Zhanpeng Wu

Staff Engineer