Milvus Manager is a web-based visual management tool for Vector Retrieval Service for Milvus. It provides a unified console to manage database objects, perform data operations, debug vector searches, configure users and permissions, and monitor instance status.
Milvus Manager integrates with Alibaba Cloud services such as Data Lake Formation (DLF) and Resource Access Management (RAM) to simplify development, debugging, and operations.
Features
Milvus Manager provides the following capabilities:
Data and object management
-
Visual schema management: Create, view, and modify collection schemas through a graphical interface. View field types, indexes, default values, and descriptions at a glance. Dynamic schema and consistency level configuration are also supported.
-
Lifecycle management for databases and collections: Create, view, edit, load, release, and delete databases and collections from a single interface.
-
Simplified vector lake management: Create and manage external collections connected to DLF lake tables to build vector lake solutions.
-
Data import, export, and editing: Filter and query data by expression, primary key, or vector. View data, perform row-level edits, and import small batches from CSV or JSON files.
Search and debugging
-
Visual vector search: Input or upload query vectors, configure search parameters, and view results with distance metrics in real time to debug and optimize search performance.
-
Expression queries and hybrid search: Combine scalar filtering expressions, primary key queries, and hybrid searches across vector and scalar fields to locate target data.
Permissions, security, and observability
-
Visual RBAC permission management: Manage users, roles, and privilege groups. Nine built-in privilege groups grant ReadOnly, ReadWrite, and Admin access at the cluster, database, and collection levels. Custom authorization and permission reuse are supported, with RAM integration for centralized access management across cloud resources such as DLF.
-
Segment-level observability and operations: View the distribution, status, and statistics of a collection's segments. Manually trigger compaction and flush for advanced tuning and troubleshooting.
Prerequisites
Before you use Milvus Manager, make sure you meet the following requirements:
-
You have an activated Alibaba Cloud account.
-
You have created a Milvus instance. For more information, see Quick start: Create a Milvus instance.
-
You have credentials for at least one database user in the Milvus instance. The default initial user is
root.
References
-
To connect to a Milvus instance from the console, see Connect to Milvus with Milvus Manager.
-
To manage databases and collections, see Manage databases.
-
To configure users, roles, and privilege groups, see Manage users, roles, and permission groups.
-
To perform vector search on DLF lake tables, see Manage external collections.
-
To view frequently asked questions, see FAQ.
Notes
-
Available features and operations may vary depending on the Milvus version, instance specifications, or permission scope. Refer to the console for the actual options.
-
Some data import, permission management, and monitoring capabilities may be restricted by the instance status or role permissions.