ClickHouse at 10: the analytics database that built itself in the open
ClickHouse marked a decade as open source on June 15, 2026, having grown into the most widely used open-source analytical database with over 2,000 contributors. Creator Alexey Milovidov frames the project around a tiered view of openness — from code that’s merely readable, up to fully transparent development with public roadmaps, CI, code review, and contributor support — and argues ClickHouse aims for the top tier deliberately. The goal is for the codebase to double as a teaching artifact: a reference for building a database, a showcase of modern C++ and the unglamorous machinery of build systems and testing, and a sandbox where experimental data structures and performance tricks get production-grade scrutiny even when they’re never meant to merge. Contributors are credited in the changelog and inside the database itself, and maintainers will finish or rewrite incomplete patches while preserving original authorship.
The technical origin story predates the open-source release by years. ClickHouse grew out of Milovidov’s work on a Google Analytics-style web analytics system backed by MySQL and hand-rolled C++ data structures, where ballooning log volumes and real-time ingestion constantly threatened to fall behind. After testing and discarding a string of existing tools — TokuDB, Hadoop, Vertica, MonetDB, Infobright, and others that couldn’t handle roughly 100 billion rows a day across 500 columns — he built two custom engines: a column-oriented store (OLAPServer, 2008) for fast on-the-fly aggregation of non-aggregated logs, and a row-oriented, CRDT-based structure (Metrage) for real-time incremental aggregation with background merges.
Those two prototypes covered the company’s needs for years, but the conceptual leap that became ClickHouse was the idea of fusing them — combining columnar storage for aggregation speed with a merge-tree design that supports real-time updates and data locality. That synthesis is the lineage behind the MergeTree engine at the heart of the modern system, and the post positions ClickHouse’s open development practices as inseparable from its engineering.
Read the full article
Continue reading at Hacker News →This is an AI-generated summary. Read the original for the full story.