Mooncake-Labs/pg_mooncake

Health v2 · maintenance only

Adoption not part of the score

2003 stars · 71 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

pg_mooncake is a Postgres extension that maintains a columnstore mirror of Postgres tables in Apache Iceberg, enabling fast real-time analytics with sub-second freshness. It uses moonlink for real-time ingestion and DuckDB for query acceleration, while keeping data accessible to other query engines.

Use cases

• run fast analytical queries on Postgres tables
• create a columnstore mirror of Postgres data in Iceberg
• real-time analytics with sub-second freshness on Postgres
• query lakehouse data from Postgres
• make Postgres data accessible to other query engines via Iceberg
• accelerate OLAP workloads on an existing Postgres database

When to choose

• you already run Postgres and need fast analytical queries without moving to a separate warehouse
• you want Iceberg-native columnstore data that other engines can also read
• you need real-time ingestion of INSERT/UPDATE/DELETE with sub-second freshness
• you want DuckDB-accelerated analytics inside Postgres

When to avoid

• you only need simple OLTP workloads with no analytical queries
• you cannot enable logical replication or preload extensions on your Postgres instance
• you need a fully managed cloud warehouse rather than a self-hosted extension
• you run a Postgres version outside the supported 14-18 range

Facets

plugin · maturity active

database analytics search-engine etl databases analytics big-data self-hosted rust cpp postgres-extension columnstore iceberg lakehouse duckdb parquet delta-lake real-time-analytics data-engineering linux macos docker

Member repositories

For agents

markdown · JSON · MCP: product_card(name="Mooncake-Labs/pg_mooncake")