Milvus Vector ETL Target

Milvus ETL Target allows you to emit and manage data from your Gathr application to Milvus.

Target Configuration

Configure the data emitter parameters as explained below.


Connection Name

A connection name can be selected from the list if you have created and saved connection details for OpenAI earlier. Or create one as explained in the topic - Milvus Connection →


Username

Provide the username with access to Milvus.

Password

Provide the password to authenticate the Milvus user.

Database Name

The name of the database to be connected. If left blank, the ‘Default’ database will be used for connection.


Collection

Select the collection of Milvus to write the vector data into it.


Collection Info

Additional information about the selected collection will be provided in the collection info section.


Create Collection

Select the checkbox to create a new collection. A new Index should also be created with a new Collection.

Collection Name

Provide collection name to be created.

Number of Shards

Number of shards for the collection to create.

Description

Provide a brief description about the collection to be created.


Index

Select the index of Milvus to write the vector data into it.


Index Info

Additional information about the selected index will be provided in the index info section.


Create Index

Select the checkbox to create a new index of the Milvus Vector.

Index Name

Provide index name to be created.

Index Type

Type of Milvus Index. Currently supported index types are: FLAT and HNSW.

Metric

Select the distance metric that will be used to measure the degree of similarity between two vectors. Supported distance metrics are L2, IP, COSINE.

Extra Param

Provide any extra parameters for creating Milvus Vector Index.

Vector Column Name

Provide the vector column name of Milvus Vector Index. This should be same as the vector Attribute specified in the Index Metadata section.


INDEX METADATA

Auto Generate ID

This option is available while creating collection. When enabled, a unique identifier is assigned to each vector automatically, ensuring that every record in the database has a distinct ID when a primary key is not explicitly defined or needed.

Output Column

Select a column name where the parsed metadata attribute should be captured.

Attribute

Select the name of the metadata attribute to be parsed.

Data Type

Select the data type of the metadata attribute being stored.

Additional Params

Provide additional parameters based on the selected data type.

Is Primary Key

This option is available while creating collection. Select the column that serves as the primary key for each record or vector in Milvus.

Auto Fill

Automatically populate fields with relevant information, streamlining the setup of index metadata in Milvus based on existing dataset characteristics.


WRITE CONFIGURATION

Save Mode

Save mode specifies how to handle any existing data in the target.

  • Append: Input data will be appended to the existing index.

  • Upsert: Contents will be inserted or updated depending on primary key values.

  • Truncate and Overwrite: This option will be enabled when user selects Create Index. This will first truncate the given collection (if collection is present), recreate the collection and then insert the data in the newly created collection.


Batch Size

Provide the batch size between 1 - 100 to determine the number of rows to insert per request.


Add Configuration: Additional properties can be added using this option as key-value pairs.


Notes

Optionally, enter notes and save the configuration.

Top