Databricks

Interact with Databricks API

Actions35

Genie Actions
Databricks SQL Actions
- Execute Query
- List Tables
Unity Catalog Actions
Model Serving Actions
Files Actions
Vector Search Actions

Overview

The node provides integration with Databricks, specifically enabling interaction with the Model Serving feature of the Databricks platform. The "Query Endpoint" operation allows users to send input data to a deployed machine learning model endpoint for inference. This is useful in scenarios where you want to automate predictions or analyses by invoking a model hosted on Databricks directly from an n8n workflow.

Practical examples include:

Sending customer data to a fraud detection model and receiving risk scores.
Querying a recommendation engine with user preferences to get personalized suggestions.
Running real-time image or text classification by sending appropriate input data to the model endpoint.

Properties

Name	Meaning
Endpoint Name	Name of the serving endpoint to which the input data will be sent for model inference.
Input Data	JSON-formatted input data that the model expects for making predictions or inferences.

Output

The node outputs the response from the model serving endpoint in the json output field. This typically contains the inference results returned by the model, such as prediction values, confidence scores, or any structured data the model produces.

If the model serving endpoint returns binary data (e.g., images or files), the node would handle it accordingly, but based on the provided properties and code, the primary output is JSON data representing the inference result.

Dependencies

Requires an active Databricks account with Model Serving enabled.
Needs proper API authentication credentials (an API token) configured in n8n to authorize requests to the Databricks API.
The node depends on the Databricks REST API endpoints for Model Serving.

Troubleshooting

Common issues:
- Incorrect or missing endpoint name can cause the request to fail.
- Malformed JSON input data may lead to errors or unexpected responses.
- Authentication failures if the API token is invalid or expired.
- Network connectivity issues between n8n and Databricks API.
Error messages and resolutions:
- 401 Unauthorized: Check that the API token credential is correctly set and has necessary permissions.
- 404 Not Found: Verify the endpoint name exists and is correctly spelled.
- 400 Bad Request: Ensure the input JSON matches the expected schema required by the model.
- Timeouts or network errors: Confirm network access and Databricks service availability.