Overview Quickstart CLI Commands .sql Models .py Models profiles.yml sources.yml Execution Paths

BUCKET STORAGES

DVT supports 4 bucket storage connectors — cloud and local. Use them as extraction sources or materialization targets. Data moves via Sling, compute runs in DuckDB.

Cloud SDKs are installed automatically by dvt sync.

HOW IT WORKS

AS A SOURCE (EXTRACTION)

S3 / GCS / Azure
       │
       │  Sling extracts files (CSV, Parquet, JSON, Delta...)
       ▼
  DuckDB Cache (.dvt/cache.duckdb)
       │
       │  Model SQL executes (DuckDB dialect)
       ▼
  Target Database (Snowflake, Postgres, etc.)

AS A TARGET (MATERIALIZATION)

Default Target (Snowflake, Postgres, etc.)
       │
       │  Model SQL executes on default target
       ▼
  Sling streams result → S3 / GCS / Azure
       │
       ▼
  Files written in configured format
  (override with config(format='...') per model)

SUPPORTED FILE FORMATS

Set a default format in profiles.yml with the format field. Override per model with config(format='delta'). Sling auto-detects format from file extension when reading.

CONNECTORS

AMAZON S3

Amazon Simple Storage Service. The most widely used cloud object store, ideal for data lakes, backups, and cross-region data distribution.

GOOGLE CLOUD STORAGE

Google's unified object storage. Tightly integrated with BigQuery, Dataflow, and the GCP ecosystem.

AZURE BLOB STORAGE

Microsoft's cloud object storage. Integrates with Fabric, Synapse, and the Azure data ecosystem.

LOCAL FILESYSTEM

Read and write files directly from the local filesystem. Ideal for development, testing, and on-premises data pipelines where data lives on shared mounts or NAS.