Milvus Vector Database: Unlocking the Future of Open Source Vector Databases
summary
Milvus is an Apache-2 open source vector database built for billion-scale, claiming real-time reads and writes with a distributed microservice architecture: proxy, query nodes, index nodes, data nodes, and object storage (S3). Ingestion uses a Kafka/Pulsar log broker, sealed segments go to an index cluster (HNSW, quantization/DISN options) and query nodes load indexes for searches. Offers embedded, standalone, and cluster deployments, targets RAG, recommendations and molecular/reverse image search, and reports subsecond eventual consistency and ~100–200 ms for strong consistency.
AI-written summary · about
chapters (11)
- 0:00Introduction
- 2:27Vectors for unstructured data
- 6:02Vector database uses
- 7:46Why Milvus stands out
- 10:59Milvus 1.0 to 2.0 evolution
- 15:57Milvus 2.0 distributed benefits
- 18:28Milvus 2.0 architecture
- 19:54Future of vector search
- 21:39Milvus as AIML infrastructure
- 23:04Choosing Milvus deployment
- 24:29Q&A on Kafka & Pulsar
related talks
37mMilvus: Scaling Vector Data Solutions with Gen AI
36mMilvus: Scaling Vector Data Solutions with Gen AI
35mMilvus 2.0: Building the World's Most Advanced Open-source Vector Database- 37mMilvus: Scaling Vector Data Solutions with Gen AI
29mThe Evolution of Milvus: A Cloud-Native Vector Database
31mMilvus 2.0: A Vector Database With Cloud-Native Architecture
51mUnleashing the Potential of Cloud-Native Vector Databases
26mThe Rise of Vector Databases - Lessons from the Milvus Community