Inserts or updates a vector in a Milvus collection. The input data is expected to be a float vector in JSON format (the dimension of the input must match dimension of the collection). Usually used together with EmbedData processor providing the float vector as the input for InsertToMilvus.
milvus, vector, vectordb, vectorstore, insert, put, update, insertai, artificial intelligence, ml, machine learning, text, LLM
In the list below, the names of required properties appear in bold. Any other properties (not in bold) are considered optional. The table also indicates any default values, and whether a property supports the NiFi Expression Language.
| Display Name | API Name | Default Value | Description |
|---|---|---|---|
| Input Format | Input Format | Single Embedding | The format of the input data. It can be either a single embedding content or a JSON document. If the input format is a single embedding content, the flow file content is expected to be a float vector in JSON format. If the input format is a JSON document, the flow file content is expected to be either a single JSON object or an array of JSON objects. |
| Vector Field Name | Vector Field Name | vector | Name of the vector field in the Milvus collection schema. Needs to be defined if the input FlowFile contains a single embedding. Otherwise the vector field name is expected to be listed in the Fields From JSON Objects property. Supports Expression Language: true (will be evaluated using flow file attributes and Environment variables) |
| Fields From Attributes | Fields From Attributes | Field names or field-attribute name pairs to insert in Milvus. The format is a comma-separated list of field_name:attribute_name pairs or field names if the attribute name is the same. The field name is the field name in the Milvus collection and the attribute name is expected to be a FlowFile attribute. The value of the FlowFile attribute will be inserted in the field. Supports Expression Language: true (will be evaluated using flow file attributes and Environment variables) | |
| Fields From JSON Objects | Fields From JSON Objects | Field names or field-JSON key pairs to insert in Milvus. At least the vector field name must be defined if the Input Format is JSON Document. The format is a comma-separated list of field_name:json_key pairs or field names if the JSON key is the same. The field name is the field name in the Milvus collection and the JSON key is expected to be a key in every JSON object in the input. The JSON value associated with the specified key will be inserted in the field. Supports Expression Language: true (will be evaluated using flow file attributes and Environment variables) | |
| Partition Name | Partition Name | Name of the partition to insert or update in Milvus. Supports Expression Language: true (will be evaluated using flow file attributes and Environment variables) | |
| Collection Name | Collection Name | Name of the collection to query in Milvus. Supports Expression Language: true (will be evaluated using flow file attributes and Environment variables) | |
| URI | URI | The URI of the Milvus instance. | |
| User | User | A valid username used to connect to the specified Milvus instance. | |
| Password | Password | A valid password used to connect to the specified Milvus instance. Sensitive Property: true | |
| DB Name | DB Name | default | The name of the database to which the target Milvus instance belongs. |
| Token | Token | A valid access token to access the specified Milvus instance. Sensitive Property: true |