Actions35
- Genie Actions
- Databricks SQL Actions
- Unity Catalog Actions
- Model Serving Actions
- Files Actions
- Vector Search Actions
Overview
The node provides integration with a vector search service within the Databricks platform, specifically allowing users to upsert (insert or update) data into a vector search index. This operation is useful for maintaining and updating searchable vector representations of data, which can be leveraged in applications such as semantic search, recommendation systems, or similarity matching.
Typical use cases include:
- Adding new vector embeddings to an existing index.
- Updating existing entries in a vector search index to reflect changes in the underlying data.
- Keeping vector indexes synchronized with source data for real-time or batch processing scenarios.
For example, a user might upsert customer profile vectors to a vector index to enable fast similarity searches for personalized recommendations.
Properties
| Name | Meaning |
|---|---|
| Index Name | The name of the vector search index where the data will be upserted. |
| Data | The JSON-formatted data representing the vectors and associated metadata to upsert. |
Output
The node outputs JSON data reflecting the result of the upsert operation. This typically includes confirmation of successful insertion or update of the provided vectors in the specified index. The exact structure depends on the API response but generally contains status information and any relevant metadata about the upserted entries.
No binary data output is indicated for this operation.
Dependencies
- Requires an active connection to the Databricks platform with appropriate API credentials (an API key or token).
- The node expects the Databricks host URL and authentication token to be configured in the node credentials.
- The vector search feature must be enabled and accessible in the Databricks environment.
Troubleshooting
- Invalid Index Name: If the specified index does not exist or is misspelled, the operation may fail. Verify the index name before running the node.
- Malformed Data: The "Data" property must be valid JSON formatted according to the vector search API requirements. Invalid JSON or incorrect schema will cause errors.
- Authentication Errors: Ensure that the API token or credentials are valid and have sufficient permissions to perform upsert operations.
- Network Issues: Connectivity problems to the Databricks host can cause timeouts or failures. Check network access and endpoint availability.
Common error messages might include authorization failures, invalid request payloads, or resource not found errors. Reviewing the error details returned by the API will guide corrective actions.