Databricks Connection
Description
Databricks Connection stores workspace credentials for the Databricks Jobs / workspace REST API. It is not the same as the Databricks JDBC relational connection (SQL warehouse).
Use this metadata type when orchestrating jobs from Hop (for example the upcoming Databricks Job Run workflow action) or when testing Jobs API access from the connection editor.
Options
| Option | Description |
|---|---|
Name | Hop metadata name (how actions reference this connection). |
Description | Optional free-text description. |
Workspace host | Workspace URL host, e.g. |
Personal access token | Databricks PAT (stored as a password field). Prefer Hop variables / secret backends for production. Tokens are never written to Hop logs by the Jobs client. |
API base path | Jobs API prefix. Default: |
Test connection
Use Test connection in the editor to call the workspace identity endpoint (SCIM Me) or fall back to listing jobs. A successful test shows the resolved user name or host.
VFS
For Hop file dialogs and VFS-aware transforms, create a separate Databricks VFS Connection under File storage. That object’s name is the URI scheme; it references this connection for host and PAT.
Example: Jobs connection prod-workspace, VFS connection name dbx-jars:
dbx-jars:///Volumes/apache-hop/default/jars/data.csv
Related
-
Databricks VFS Connection — named scheme for Volumes / Workspace
-
Native Spark on Databricks — Deploy & run, classic clusters, Volumes
-
Databricks Job Run — trigger and wait for Jobs API runs
-
Databricks database (JDBC) — SQL warehouse access
-
Native Spark pipeline engine — fat jar / MainSpark on cluster
-
Spark Catalog — Unity Catalog templates for lakehouse TABLE mode