Actions42
- Genie Actions
- Databricks SQL Actions
- Unity Catalog Actions
- Model Serving Actions
- Files Actions
- Vector Search Actions
Overview
This node integrates with the Databricks API, enabling users to interact with various Databricks services including Model Serving. Specifically, for the Model Serving resource and the Create Serving Endpoint operation, it allows users to create a new model serving endpoint by specifying which models to serve and how traffic should be routed among them.
This functionality is beneficial in scenarios where machine learning models need to be deployed as scalable endpoints for real-time inference. For example, data scientists or ML engineers can use this node to programmatically deploy multiple models behind a single endpoint and control traffic distribution for A/B testing or gradual rollouts.
Properties
| Name | Meaning |
|---|---|
| Served Models | List of models to serve, provided in JSON format. This defines which models will be hosted on the endpoint. |
| Traffic Config | Traffic configuration for the endpoint, also in JSON format. This controls how incoming requests are distributed across the served models. |
Output
The node outputs an array of JSON objects corresponding to each input item processed. For the Create Serving Endpoint operation, the output JSON typically contains the response from the Databricks API confirming the creation or update of the serving endpoint, including details such as endpoint ID, status, and configuration.
No binary data output is indicated for this operation.
Dependencies
- Requires an API authentication token credential to access the Databricks API.
- The node expects the Databricks host URL and token to be configured in the credentials.
- The user must provide valid JSON strings for the "Served Models" and "Traffic Config" properties that conform to the Databricks API schema for model serving endpoints.
Troubleshooting
- Invalid JSON Format: If the "Served Models" or "Traffic Config" inputs are not valid JSON, the API request will fail. Ensure these fields contain properly formatted JSON.
- Authentication Errors: Missing or incorrect API tokens will cause authorization failures. Verify that the API key/token is correctly set up in the node credentials.
- API Errors: The node logs detailed error messages if the Databricks API returns errors (e.g., invalid parameters, permission issues). Review the error details in the node execution logs.
- Network Issues: If no response is received from the server, check network connectivity and firewall settings.
- Continue On Fail: If enabled, the node will continue processing subsequent items even if some fail, returning error details in the output JSON.