Query Load Balancing

Milvus 2.0 architecture enables real-time data query by orchestrating data loading, management, and query execution across various components. QueryNode now supports plugins, expanding querying scenarios. The compaction algorithm has been improved for faster queries. The default number of collection shards has been reduced to one for better resource saving and querying performance. An improved query grouping algorithm increases Milvus's throughput by at least 20%. Milvus 2.2.8 also addresses red. Milvus's query coordinator distributes segments evenly to each query node according to RAM usage, triggering load balancing when memory usage exceeds thresholds or the difference between nodes is too large. Segments are transferred, and the source node releases them after successful loading on the destination.