pg_duckdb: DuckDB-powered Postgres

on YouTube ↗03m

" summary

pg_duckdb shows a Postgres setup that exports a hits table with 100 million rows and joins a users table with 17.6 million users by copying Parquet to S3, caching it via a secrets entry in the duckdb_secrets table, and exposing Postgres views that read from S3. A simple count that used to take five minutes now reads the hits.par view in 300 milliseconds, a top-users query finishes in 900 milliseconds, and results copy to S3 as CSV in 5 seconds.

AI-written summary · about

" chapters (3)

1. 0:00Introduction
2. 0:58Postgres views
3. 1:31Running the query

" related talks

• 24mpg_duckdb: Ducking awesome analytics in PostgresJelte Fennema-Nio · POSETTE: An Event for Postgres 2025
• 21mWhat can Postgres learn from DuckDB?Jelte Fennema-Nio · PGConf.dev 2025
• 19mIn-Process Analytical Data Management with DuckDBHannes Mühleisen · posit::conf(2023)
• 01h 02mAnalytics for not-so-big data with DuckDBDavid Ostrovsky · NDC Oslo 2025
• 24mpg_duckdb in Action: Accelerating Analytics on Azure Database for PostgreSQLNitin Jadhav · POSETTE: An Event for Postgres 2026
• 27mThe Duck and the DataFrame: A Data Engineer’s Journey with DuckDBAnkur Jain · PyCon AU 2025
• 33mData preparation and analytics on your laptop with DuckDBMatson, Prawiroatmodjo · PyData NYC 2024
• 40mDuckDB: Harnessing in-process analytics for data science and beyondGabor Szarnyas · Øredev 2023