Skip to content
Search

ClickHouse

lastmod 2025-06-09created 2021-05-3111 edits

Open-source column-based OLAP DB

Features

  • RDBMS for Analytics
  • Fast, near-real-time speed, scalable, reliable
    • 5x–24x faster than existing column DBs (Vertica, Greenplum)
    • 100x–1000x faster than Hive or MySQL
  • Easy setup
  • Rich features : distributed/parallel query execution, real-time query processing, data compression, Hot/Cold storage separation, querying JSON documents, distributed Join, S3-compatible object storage..
  • Fault-tolerant : multi-master asynchronous replication
  • Highly Reliable : distributed system with no SPOF
  • Makes the most of the hardware : makes full use of SIMD CPU instructions, minimizes drive access/data transfer, etc.

One of 7 databases for 2025

  • See Database#7 databases for 2025 (GeekNews)
  • ClickHouse is a database specialized for OLAP workloads
    • The combination of PostgreSQL for OLTP and ClickHouse for OLAP is ideal
    • Handles large-scale analytics workloads and supports high data insertion rates through horizontal scaling and sharding
  • Key features of ClickHouse
    • Tiered storage support:
      • Can store "hot data" and "cold data" separately
      • e.g.: GitLab's documentation covers a use case of this in detail
    • Large dataset processing and real-time analytics:
      • Suitable for datasets too large for DuckDB to handle
      • Delivers strong performance where real-time analytics is needed
  • Ease of operation
    • Operations documentation on deployment, scaling, backup, etc. is well organized and detailed
    • e.g.: Provides documentation that even covers how to configure the CPU properly
  • Learning ClickHouse is recommended
    • Experiment with large analytics datasets, or convert analyses done with DuckDB to ClickHouse
    • chDB, the embedded version of ClickHouse, allows a more direct comparison with SQLite

See also

Favorite site

author Zer0 sitemap