How to use Milvus Vector Database with scale and multi-in-memory replicate of a collection

If you did not know what a Milvus Vector database is and how it’s been set up into an Amazon EKS (Elastic Kubernetes Service) cluster you can read my article Run a Milvus vector database inside your Amazon EKS (Elastic Kubernetes Service) in cluster mode to learn more about Milvus.

In this article, I will explain why you should use more than only 1 single query node in your Milvus cluster and how you can manually scale more than one Milvus Query Node inside your Kubernetes cluster.

Benefits of using multi Milvus Query Nodes in your Milvus Cluster

If you are using Milvus with only one Query Node its means if this node crashing its takes a long time to restore and you can search in the Milvus Vector database until it's restored again.

So I will recommend running 3 or more Milvus Query Nodes in your cluster to reduce the risk of one point of failure and adding the high availability architecture to your Milvus cluster.