@kapa.ai difference between airbyte oss connector docker images and pyairbyte?

Here's a breakdown of the key differences between Airbyte OSS connector Docker images and PyAirbyte: Airbyte OSS Connector Docker Images • Connectors are packaged as Docker images and run via the Airbyte platform. • They follow the Airbyte Protocol, accepting commands like

spec

,

check

,

discover

, and `read`/`write` via Docker CLI. [Airbyte Protocol Docker] • The Docker image must meet specific requirements: an

AIRBYTE_ENTRYPOINT

environment variable with an absolute path, a non-root

airbyte

user, and a

/airbyte

directory. [Airbyte Protocol Docker] • Connector metadata (Docker repository, image tag, registry inclusion) is managed via a

metadata.yaml

file, which determines whether a connector appears in the OSS or Cloud registry. [Connector Metadata.yaml] • Designed for batch ELT/ETL pipelines orchestrated by the Airbyte platform, with schedules and state tracking for large-volume data replication into warehouses or data lakes. PyAirbyte • PyAirbyte is a Python library that lets you use Airbyte connectors directly in Python code, without necessarily running the full Airbyte platform. • It supports multiple execution methods for connectors: Python-based (pip), YAML manifest, or Docker image — selected automatically based on connector metadata or explicitly via parameters like

docker_image=True

. [PyAirbyte experimental] • Most destination connectors are Java-based and only available as Docker images; if Docker is unavailable, PyAirbyte offers SQL caches as an alternative. [PyAirbyte destinations] • Suited for developer environments, pulling data into SQL caches (DuckDB, Postgres, BigQuery, etc.) or writing scripts for data pipelines. Summary In short, OSS connector Docker images are the underlying connector artifacts run by the Airbyte platform, while PyAirbyte is a Python library that can use those same connectors (among other methods) directly in your code without requiring the full platform. 👍 0 button 👎 0 button ℹ️ Tip: Mention @kapa.ai in the thread for follow-ups or DM the app for 1-to-1 help., with interactive elements