0.1.0 release is here 🎉 Please try it out!

pg_duckdb: Official Postgres extension for DuckDB

pg_duckdb is a Postgres extension that embeds DuckDB's columnar-vectorized analytics engine and features into Postgres. We recommend using pg_duckdb to build high performance analytics and data-intensive applications.

pg_duckdb was developed in collaboration with our partners, Hydra and MotherDuck.

Features

See our official documentation for further details.

SELECT queries executed by the DuckDB engine can directly read Postgres tables. (If you only query Postgres tables you need to run SET duckdb.force_execution TO true, see the IMPORTANT section above for details)
- Able to read data types that exist in both Postgres and DuckDB. The following data types are supported: numeric, character, binary, date/time, boolean, uuid, json, and arrays.
- If DuckDB cannot support the query for any reason, execution falls back to Postgres.
Read and Write support for object storage (AWS S3, Cloudflare R2, or Google GCS):
- Read parquet, CSV and JSON files:
  - SELECT n FROM read_parquet('s3://bucket/file.parquet') AS (n int)
  - SELECT n FROM read_csv('s3://bucket/file.csv') AS (n int)
  - SELECT n FROM read_json('s3://bucket/file.json') AS (n int)
  - You can pass globs and arrays to these functions, just like in DuckDB
- Enable the DuckDB Iceberg extension using SELECT duckdb.install_extension('iceberg') and read Iceberg files with iceberg_scan.
- Enable the DuckDB Delta extension using SELECT duckdb.install_extension('delta') and read Delta files with delta_scan.
- Write a query — or an entire table — to parquet in object storage.
  - COPY (SELECT foo, bar FROM baz) TO 's3://...'
  - COPY table TO 's3://...'
  - Read and write to Parquet format in a single query
```
COPY (
    SELECT count(*), name
    FROM read_parquet('s3://bucket/file.parquet') AS (name text)
    GROUP BY name
    ORDER BY count DESC
) TO 's3://bucket/results.parquet';
```
Read and Write support for data stored in MotherDuck
Query and JOIN data in object storage/MotherDuck with Postgres tables, views, and materialized views.
Create temporary tables in DuckDB its columnar storage format using CREATE TEMP TABLE ... USING duckdb.
Install DuckDB extensions using SELECT duckdb.install_extension('extension_name');
Toggle DuckDB execution on/off with a setting:
- SET duckdb.force_execution = true|false
Cache remote object locally for faster execution using SELECT duckdb.cache('path', 'type'); where
- 'path' is HTTPFS/S3/GCS/R2 remote object
- 'type' specify remote object type: 'parquet' or 'csv'

Installation

Docker

Docker images are available on Dockerhub and are based on the official Postgres image. Use of this image is the same as the Postgres image. For example, you can run the image directly:

docker run -d -e POSTGRES_PASSWORD=duckdb pgduckdb/pgduckdb:16-main

And with MotherDuck, you only need a a MotherDuck access token and then it is as simple as:

$ export MOTHERDUCK_TOKEN=<your personal MD token>
$ docker run -d -e POSTGRES_PASSWORD=duckdb -e MOTHERDUCK_TOKEN pgduckdb/pgduckdb:16-main -c duckdb.motherduck_enabled=true

Or you can use the docker compose in this repo:

git clone https://github.com/duckdb/pg_duckdb && cd pg_duckdb && docker compose up -d

Once started, connect to the database using psql:

psql postgres://postgres:duckdb@127.0.0.1:5432/postgres
# Or if using docker compose
docker compose exec db psql

For other usages see our Docker specific README.

pgxman (apt)

Pre-built apt binaries are available via pgxman. After installation, you will need to add pg_duckdb to shared_preload_libraries and create the extension.

pgxman install pg_duckdb

Note: due to the use of shared_preload_libraries, pgxman's container support is not currently compatible with pg_duckdb.

Compile from source

To build pg_duckdb, you need:

Postgres 15-17
Ubuntu 22.04-24.04 or MacOS
Standard set of build tools for building Postgres extensions
Build tools that are required to build DuckDB

To build and install, run:

make install

Add pg_duckdb to the shared_preload_libraries in your postgresql.conf file:

shared_preload_libraries = 'pg_duckdb'

Next, create the pg_duckdb extension:

CREATE EXTENSION pg_duckdb;

IMPORTANT: DuckDB execution is usually enabled automatically when needed. It's enabled whenever you use DuckDB functions (such as read_csv), when you query DuckDB tables, and when running COPY table TO 's3://...'. However, if you want queries which only touch Postgres tables to use DuckDB execution you need to run SET duckdb.force_execution TO true'. This feature is opt-in to avoid breaking existing queries. To avoid doing that for every session, you can configure it for a certain user by doing ALTER USER my_analytics_user SET duckdb.force_execution TO true.

Getting Started

See our official documentation for more usage information.

pg_duckdb relies on DuckDB's vectorized execution engine to read and write data to object storage bucket (AWS S3, Cloudflare R2, or Google GCS) and/or MotherDuck. The follow two sections describe how to get started with these destinations.

Object storage bucket (AWS S3, Cloudflare R2, or Google GCS)

Querying data stored in Parquet, CSV, JSON, Iceberg and Delta format can be done with read_parquet, read_csv, read_json, iceberg_scan and delta_scan respectively.

Add a credential to enable DuckDB's httpfs support.

-- Session Token is Optional
INSERT INTO duckdb.secrets
(type, key_id, secret, session_token, region)
VALUES ('S3', 'access_key_id', 'secret_access_key', 'session_token', 'us-east-1');

Copy data directly to your bucket - no ETL pipeline!

COPY (SELECT user_id, item_id, price, purchased_at FROM purchases)
TO 's3://your-bucket/purchases.parquet;

Perform analytics on your data.

SELECT SUM(price) AS total, item_id
FROM read_parquet('s3://your-bucket/purchases.parquet')
  AS (price float, item_id int)
GROUP BY item_id
ORDER BY total DESC
LIMIT 100;

Connect with MotherDuck

pg_duckdb also integrates with MotherDuck. To enable this support you first need to generate an access token and then add the following line to your postgresql.conf file:

duckdb.motherduck_token = 'your_access_token'

NOTE: If you don't want to store the token in your postgresql.conffile can also store the token in the motherduck_token environment variable and then explicitly enable MotherDuck support in your postgresql.conf file:

duckdb.motherduck_enabled = true

If you installed pg_duckdb in a different Postgres database than the default one named postgres, then you also need to add the following line to your postgresql.conf file:

duckdb.motherduck_postgres_database = 'your_database_name'

After doing this (and possibly restarting Postgres). You can then you create tables in the MotherDuck database by using the duckdb Table Access Method like this:

CREATE TABLE orders(id bigint, item text, price NUMERIC(10, 2)) USING duckdb;
CREATE TABLE users_md_copy USING duckdb AS SELECT * FROM users;

Any tables that you already had in MotherDuck are automatically available in Postgres. Since DuckDB and MotherDuck allow accessing multiple databases from a single connection and Postgres does not, we map database+schema in DuckDB to a schema name in Postgres.

This is done in the following way:

Each schema in your default MotherDuck database are simply merged with the Postgres schemas with the same name.
Except for the main DuckDB schema in your default database, which is merged with the Postgres public schema.
Tables in other databases are put into dedicated DuckDB-only schemas. These schemas are of the form ddb$<duckdb_db_name>$<duckdb_schema_name> (including the literal $ characters).
Except for the main schema in those other databases. That schema should be accessed using the shorter name ddb$<db_name> instead.

An example of each of these cases is shown below:

INSERT INTO my_table VALUES (1, 'abc'); -- inserts into my_db.main.my_table
INSERT INTO your_schema.tab1 VALUES (1, 'abc'); -- inserts into my_db.your_schema.tab1
SELECT COUNT(*) FROM ddb$my_shared_db.aggregated_order_data; -- reads from my_shared_db.main.aggregated_order_data
SELECT COUNT(*) FROM ddb$sample_data$hn.hacker_news; -- reads from sample_data.hn.hacker_news

Roadmap

Please see the project milestones for upcoming planned tasks and features.

Contributing

pg_duckdb was developed in collaboration with our partners, Hydra and MotherDuck. We look forward to their continued contributions and leadership.

Hydra is a Y Combinator-backed database company, focused on DuckDB-Powered Postgres for app developers.

duckdb / pg_duckdb

readme