Case Study: Lemit achieves 100K+ QPS on MySQL with Readyset

Lemit reaches 109K QPS on MySQL with Readyset

Lemit, a Brazilian data enrichment and credit intelligence company, faced a critical challenge scaling its MySQL database to handle its high-volume, latency-sensitive workload. Despite aggressive vertical scaling to massive hardware and using read replicas, read queries were still competing with writes, causing throughput to plateau. The mature application codebase had no room for changes, and they needed a solution that could break through the performance ceiling without any modifications.

The vendor ReadySet provided a SQL-level caching solution that was deployed as a transparent, wire-compatible layer in front of MySQL. By automatically caching results and keeping them fresh via the MySQL binlog, ReadySet delivered sub-millisecond latency for cached queries. This implementation required zero code changes and immediately allowed Lemit to achieve over 109,000 queries per second, an 8x increase in read throughput, while freeing the primary database to handle writes.

Lemit