dbt-redis-adbc

A community dbt adapter for Redis 8.x using the Redis Arrow Database Connectivity (ADBC) driver. Supports Python-based dbt 1.x; dbt 2.x and the dbt platform are not supported.

Install

Requires Python 3.10 or later and Redis 8.x with the Query Engine (Search) enabled. The adapter is tested against Redis 8.6.2.

python -m pip install dbt-redis-adbc

The Python package does not include the native Redis ADBC driver. Download and extract the shared library for your platform from the driver releases. The example project uses driver v0.0.14. Prebuilt libraries are available for macOS Apple Silicon, Linux x86-64, and Linux arm64.

Set REDIS_ADBC_DRIVER to the library's absolute path. See the driver's server requirements and ACL permissions before connecting.

Configure

Set your dbt project's profile to redis_project and add this profile to ~/.dbt/profiles.yml:

redis_project:
  target: dev
  outputs:
    dev:
      type: redis_adbc
      driver: "{{ env_var('REDIS_ADBC_DRIVER') }}"
      uri: "{{ env_var('REDIS_URI', 'redis://localhost:6379/0') }}"
      username: "{{ env_var('REDIS_USERNAME', '') }}"
      password: "{{ env_var('REDIS_PASSWORD', '') }}"
      database: redis
      schema: analytics
      threads: 4

Use a rediss:// URI for TLS. Keep credentials in the separate username and password fields. From your project, verify the connection with:

dbt debug --profile redis_project

Features and limitations

• Tables, views, ephemeral models, Arrow bulk-loaded seeds, and snapshots
• SQL incremental models with append, delete+insert, merge, and microbatch
• Model contracts, persisted relation and column descriptions, and Python table models
• No SQL transactions or rollback: statements autocommit, and interrupted writes may need cleanup or a rerun
• No materialized views or SQL grants; configure access with Redis ACLs
• Python models run in the dbt process and load their inputs into its memory; incremental Python models are not supported

See the adapter reference for all profile options, supported features, and limitations. The example project and functional tests demonstrate the adapter with NYC taxi data.