Zilliz Vector Store icon

Zilliz Vector Store

Work with Zilliz vector database for AI applications

Actions2

Overview

This node integrates with the Zilliz vector database, enabling AI applications to perform vector similarity searches and manage collections. It is particularly useful for scenarios involving semantic search, recommendation systems, or any application requiring efficient nearest neighbor search in high-dimensional vector spaces.

A common use case is searching a vector collection to find items most similar to a given query vector, optionally filtered by metadata conditions. For example, you might search an image embedding database to find visually similar images or query a text embedding collection to retrieve semantically related documents.

Properties

Name Meaning
Database Name Name of the Zilliz database to use. Defaults to "default" if not specified.
Collection Name Name of the Zilliz collection within the database to search. Must start with a letter and contain only letters, numbers, and underscores.
Query Vector The vector (array of numbers) to search for similarity against stored vectors in the collection. Must be valid JSON array.
Limit Maximum number of search results to return. Minimum value is 1. Default is 50.
Filter Expression Optional filter expression to apply on the search results, e.g., "ID > 100".

Output

The node outputs an array of JSON objects representing the search results:

  • For the Search Vectors operation, the output JSON contains a results field which is an array of matched vectors with their associated metadata and similarity scores.
  • For the List Collections operation, the output JSON contains a collections field listing all collections available in the specified database.

No binary data output is produced by this node.

Example output structure for a search operation:

{
  "results": [
    {
      "id": "vector_id_1",
      "score": 0.95,
      "metadata": { /* additional fields */ }
    },
    {
      "id": "vector_id_2",
      "score": 0.93,
      "metadata": { /* additional fields */ }
    }
  ]
}

Dependencies

  • Requires an API key credential and cluster endpoint configuration for authenticating with the Zilliz vector database service.
  • Uses the ZillizClient class from a shared library to interact with the Zilliz API.
  • The node expects the user to provide valid JSON for the query vector input.

Troubleshooting

  • Invalid JSON in Query Vector parameter: This error occurs if the provided query vector is not valid JSON or not an array. Ensure the vector is a properly formatted JSON array of numbers, e.g., [0.1, 0.2, 0.3].
  • Collection name validation: Collection names must start with a letter and contain only letters, numbers, and underscores. Invalid names will cause errors.
  • Unknown operation: If an unsupported operation is selected, the node will throw an error. Only "search" and "listCollections" are supported.
  • API authentication issues: Make sure the API key and cluster endpoint credentials are correctly configured and have sufficient permissions.
  • If the node is set to continue on fail, errors for individual items will be returned in the output JSON under an error field instead of stopping execution.

Links and References

Discussion