Scaling a 1.59 quadrillion-row logging system across three clouds
ClickHouse partnered with a major tech company to scale LogHouse, a high-volume logging system handling 1.59 quadrillion rows across AWS, GCP, and Azure. The system processes 1.2 PB of new data daily, using optimized compression, tiered storage, and real-time replication to achieve sub-second query performance at minimal cost.
Background
ClickHouse is an open-source, column-oriented database designed for real-time analytics on massive datasets. LogHouse is ClickHouse's internal logging infrastructure — a system that ingests and stores logs (machine-generated event records) from all of ClickHouse's cloud services. This blog post describes how the engineering team scaled LogHouse to handle 1.59 quadrillion rows (that's 1.59 × 10¹⁵) across three major cloud providers (AWS, GCP, and Azure), a scale only a handful of systems in the world have achieved. It covers the architectural changes — like sharding strategies, tiered storage (hot/warm/cold), and cross-cloud replication — that made this possible while keeping query performance and cost under control. For readers interested in distributed systems, observability infrastructure, or database engineering at extreme scale, this provides a concrete case study of real-world trade-offs.