Skip to main content

Overview of Google Drive ingestion setup

The managed Google Drive connector in Lakeflow Connect ingests unstructured files (such as PDFs and Google Docs) and structured formats (such as CSV, JSON, and Excel) into Delta tables, and supports three authentication methods. For custom pipelines that use Spark and SQL APIs with full control over parsing and transformations, see Ingest files from Google Drive.

Choose your Google Drive connector

Lakeflow Connect offers two Google Drive connectors. They both access data in Google Drive, but differ in their level of management.

Connector

Description

Managed Google Drive connector

A fully managed, low-maintenance connector for enterprise applications that ingests data into Delta tables and keeps them in sync with the source.

Standard Google Drive connector

Build custom ingestion pipelines with SQL, PySpark, or Lakeflow pipelines using batch and streaming APIs such as read_files, spark.read, COPY INTO, and Auto Loader. Offers the flexibility to perform complex transformations during ingestion, while giving you greater responsibility for managing and maintaining your pipelines.

Connector

Description

Managed Google Drive connector

A fully managed, low-maintenance connector for enterprise applications that ingests data into Delta tables and keeps them in sync with the source.

Standard Google Drive connector

Build custom ingestion pipelines with SQL, PySpark, or Lakeflow pipelines using batch and streaming APIs such as read_files, spark.read, COPY INTO, and Auto Loader. Offers the flexibility to perform complex transformations during ingestion, while giving you greater responsibility for managing and maintaining your pipelines.

Which authentication method should I choose?

The managed Google Drive connector supports the following authentication methods:

Both OAuth U2M methods authenticate as an individual Google account, so the connector ingests only the files that account can access. To ingest content that isn't tied to one person's account, use an OAuth service account.

If OAuth U2M: Databricks-managed isn't listed in your workspace, use OAuth U2M: Custom-managed.