What ClickHouse is for

What ClickHouse is for#

ClickHouse is optimized for workloads where speed beats everything else:

  • High-volume analytics where data arrives seconds or minutes ago and must be queryable immediately

  • IoT monitoring, ad-tech bidding, fraud detection, and similar time-critical workloads

  • User-facing dashboards with thousands of concurrent users on live data

  • Sub-second rollups on events, logs, and telemetry

  • Postgres observability, storing and querying metrics from Postgres databases at scale

ClickHouse and WarehousePG#

ClickHouse and WarehousePG answer different questions and complement each other within the EDB Postgres AI Analytics pillar.

ClickHouse answers “What’s happening right now?”: sub-second aggregations on high-volume, recent data, optimized for few-column scans across huge single tables.

WarehousePG answers “What happened?”: complex multi-table joins, enrichment, data modeling, and batch analytics where accuracy matters more than speed.

Apache Iceberg® acts as the shared source of truth between them: data written by either engine is accessible to both, with no copies or ETL pipelines between them.

Contents