Clustering is a feature that allows two or more AEM instances to share the same repository view, whether they are running on the same machine or on different machines. All cluster instances should, in theory, have the same content, and requests to any node in the cluster should yield the same result. EVEN when operating a single instance, CQ5 CRX is pre-loaded to run in a cluster. As a result, CQ5 facilitates the deployment of multi-node clusters with minimal effort. 

Clustering in CQ has the following advantages:

• High Availability - Clustering helps to ensure high availability in the event of a server or repository failure.
• Improve Performance - To improve performance, the load can be balanced between cluster nodes.
• Backup - If one of the nodes is serving as a passive node, real-time backup is feasible through clustering.

Every repository (or instance) in CQ is REFERRED to as a node. In a Cluster, there is always one master, and the other nodes are referred to as slaves. TCP/IP is used to communicate between the SLAVE and the master. If one of the instances fails, another node takes over as master and the cluster continues to FUNCTION.