Distributed DuckDB Instance

An open-source implementation of the ideas pioneered by MotherDuck — differential storage, hybrid (dual) execution, and transparent remote databases for DuckDB — available for anyone to run, extend, and build on.

MotherDuck showed that DuckDB can work beautifully in the cloud: ATTACH 'md:mydb', and remote tables appear local. Queries split transparently across your laptop and the cloud. Storage is layered and snapshot-based. OpenDuck takes those architectural ideas — differential storage, dual execution, the attach-based UX — and makes them open. Open protocol, open backend, open extension.

import duckdb
con = duckdb.connect()
con.execute("LOAD 'openduck';")
con.execute("ATTACH 'openduck:mydb?endpoint=http://localhost:7878&token=xxx' AS cloud;")
con.sql("SELECT * FROM cloud.users").show() # remote, transparent
con.sql("SELECT * FROM local.t JOIN cloud.t2 ON ...").show() # hybrid, one query

Append-only layers with PostgreSQL metadata. DuckDB sees a normal file; OpenDuck persists data as immutable sealed layers addressable from object storage. Snapshots give you consistent reads. One serialized write path, many concurrent readers.

A single query can run partly on your machine and partly on a remote worker. The gateway splits the plan, labels each operator LOCAL or REMOTE, and inserts bridge operators at the boundaries. Only intermediate results cross the wire.

[LOCAL] HashJoin(l.id = r.id)
[LOCAL] Scan(products) ← your laptop
[LOCAL] Bridge(R→L)
[REMOTE] Scan(sales) ← remote worker

The extension implements DuckDB's StorageExtension and Catalog interfaces. Remote tables are first-class catalog entries, they participate in JOINs, CTEs, and the optimizer like local tables.

OpenDuck's protocol is intentionally minimal: two RPCs defined in execution.proto. The first one to execute a query, and the other to stream results back as Arrow IPC batches.

Because the protocol is open and simple, you're not locked into a single backend. Any service that speaks gRPC and returns Arrow can serve as an OpenDuck-compatible backend. Run the included Rust gateway, replace it with your own implementation, or plug in an entirely different execution engine — the client and extension don't care what's on the other side.

MotherDuck is a commercial cloud service. OpenDuck is an open-source project inspired by its architecture. OpenDuck is not wire-compatible with MotherDuck. It reimplements the same architectural ideas as an open protocol.

Arrow Flight SQL is a generic database protocol — "JDBC/ODBC over Arrow." OpenDuck is a DuckDB-specific system with a narrower scope but deeper integration. OpenDuck's architecture draws heavily from MotherDuck's published work on differential storage, dual execution, and cloud-native DuckDB. Credit to the MotherDuck team for pioneering these ideas.

License: MIT