Case Study: Cloudflare scales Zero Trust analytics and reporting with TimescaleDB

Cloudflare boosts analytics 35x with Timescale

Cloudflare, a leading internet security and performance company, faced a challenge scaling the analytics and reporting for its Zero Trust product suite, specifically for its Digital Experience Monitoring (DEX) feature. Their initial PostgreSQL setup for handling structured log data from millions of WARP clients began to show performance degradation as tables grew to billions of rows, impacting query times for large customers. They evaluated TimescaleDB as a potential solution to improve performance without adding excessive complexity to their infrastructure.

By implementing the TimescaleDB extension, Cloudflare converted its tables to hypertables and leveraged features like automated compression, continuous aggregates, and sparse indexes. This resulted in query performance improvements of 5x to 35x and impressive compression ratios of up to 33x, allowing for significantly more data retention at the same cost. The TimescaleDB solution provided the performance of a specialized analytical database while maintaining the simplicity and PostgreSQL compatibility that Cloudflare required, leading to its adoption by other teams within the company.

Cloudflare