Crate lumosai_vector_milvus 

LumosAI Milvus Integration

This crate provides Milvus integration for LumosAI vector storage, offering high-performance vector database capabilities with cloud-native features.

Features

• High Performance: Distributed vector database optimized for large-scale applications
• Cloud Native: Kubernetes-ready with horizontal scaling
• Rich Indexing: Multiple index types (IVF, HNSW, ANNOY, etc.)
• Metadata Filtering: Complex filtering with boolean expressions
• Multi-tenancy: Collection-based isolation and resource management
• ACID Transactions: Consistency guarantees for critical operations
• Real-time: Support for real-time data ingestion and querying

Quick Start

use lumosai_vector_milvus::{MilvusStorage, MilvusConfig};
use lumosai_vector_core::traits::VectorStorage;

#[tokio::main]
async fn main() -> Result<(), Box<dyn std::error::Error>> {
    // Create Milvus storage
    let config = MilvusConfig::new("http://localhost:19530")
        .with_database("default")
        .with_auth("username", "password");
    let storage = MilvusStorage::new(config).await?;
     
    // Create a collection
    let index_config = IndexConfig::new("documents", 384)
        .with_metric(SimilarityMetric::Cosine);
    storage.create_index(index_config).await?;
     
    // Insert documents
    let docs = vec![
        Document::new("doc1", "Hello world")
            .with_embedding(vec![0.1; 384])
            .with_metadata("category", "greeting"),
    ];
    storage.upsert_documents("documents", docs).await?;
     
    Ok(())
}

Re-exports

pub use storage::MilvusStorage;

pub use config::MilvusConfig;

pub use config::MilvusConfigBuilder;

pub use error::MilvusError;

pub use error::MilvusResult;

pub use client::MilvusClient;

pub use types::MilvusEntity;

pub use types::AuthRequest;

pub use types::AuthResponse;

pub use types::CollectionInfo;

pub use types::CollectionSchema;

Modules

client

Milvus client implementation

config

Milvus configuration module

error

Error types for Milvus integration

storage

Milvus storage implementation

types

Type definitions for Milvus integration

utils

Utility functions for Milvus operations

Structs

Document

Document representation with embedding support

MilvusConnection

Milvus client for managing connections and databases

PerformanceMetrics

性能指标

Enums

MetadataValue

Metadata value that can hold various types

SimilarityMetric

Similarity metrics for vector comparison

Traits

VectorStorage

Core trait for vector storage backends

Functions

create_milvus_storage

Create a new Milvus storage instance

create_milvus_storage_with_config

Create a new Milvus storage instance with configuration