Overview¶

duckrun is glue over DuckDB + delta-rs + dbt-duckdb: a notebook connect() for SQL over Delta, and a dbt adapter that materializes models as Delta tables, on one engine split.

How it works¶

DuckDB runs every query and reads Delta through delta_scan views; delta-rs handles every write; an Arrow C-stream bridges them; dbt orchestrates on top.

duckrun architecture: DuckDB executes SQL and reads Delta via delta_scan; an Arrow C-stream bridges to delta-rs, which handles every write and commits against the read version (OCC); dbt orchestrates on top

Writes are snapshot-pinned: the read is fixed at delta_scan(…, version => N) and the write commits against N, so a concurrent commit is rejected with CommitFailedError instead of silently overwriting a lost update.

Two writers race on one table: Writer A reads v5 and computes; Writer B commits v6 in between; A's commit against v5 is rejected with CommitFailedError instead of silently overwriting B

The full model is in Snapshot isolation; the engine-split rationale in the Design document.