Use put-vector-index to create a vector index - Object Storage Service - Alibaba Cloud - Object Storage Service

Notes

A single vector bucket can contain up to 100 vector indexes.
A maximum of 5 PutVectorIndex requests are supported per second.

Permissions

By default, Alibaba Cloud accounts have all permissions, whereas Resource Access Management (RAM) users and RAM roles have no permissions. The Alibaba Cloud account or an administrator must grant permissions through a RAM policy or a bucket policy.

API	Action	Description
PutVectorIndex	`oss:PutVectorIndex`	Creates a vector index.

Command format

ossutil vectors-api put-vector-index --bucket value --index-name value [flags]

Parameters

Parameter	Type	Description
--bucket	string	The name of the vector bucket.
--data-type	string	The data type of the vector. Custom data types are not supported. Default value: float32.
--dimension	int	The vector dimensions. The value can be an integer from 1 to 4096. The default value is 512.
--distance-metric	string	The distance measure function. Valid values: euclidean: Euclidean distance (default) cosine: Cosine distance
--index-name	string	The name of the index. You can specify a custom name. The default value is empty. The name must be globally unique within the vector bucket and must be 1 to 63 characters in length. The name can contain only letters and digits, and must start with a letter.
--metadata	string	The container for metadata configuration. Only non-filterable metadata is supported. The following limits apply to the configuration of non-filterable metadata: The number of metadata entries must be from 1 to 100. The length of each metadata primary key name must be from 1 to 63 bytes.

Note

The put-vector-index command corresponds to the PutVectorIndex API operation.
For more information about the supported global command-line options, see Supported global command-line options.

Usage examples

Create an index named index for the examplebucket bucket. The index has 512 vector dimensions, uses the float32 data type, and the euclidean distance measure.
```
ossutil vectors-api put-vector-index --bucket examplebucket --index-name index --data-type float32 --dimension 512 --distance-metric euclidean
```

Create an index named index for the examplebucket bucket and configure non-filterable metadata. The index has 512 vector dimensions, uses the float32 data type, and the euclidean distance measure.

ossutil vectors-api put-vector-index --bucket examplebucket --index-name index --data-type float32 --dimension 512 --distance-metric euclidean --metadata '{"nonFilterableMetadataKeys":["owner","createdBy"]}'