What do servers do in a cluster?

The term cluster describes a computer network in which several computers or servers are combined into one unit. This is done via a network in which the individual computers, mostly high-performance servers, are connected via switches and load balancers. The individual computers in the network are referred to as nodes and can be located either centrally at one location or distributed over several locations. All nodes work independently of each other, which means that the loads in the network are distributed over several independent systems.

There is a rough distinction between two different clusters: The High Performance Computing Cluster (HPC) and the High Availability Cluster (HAC). While an HPC is designed for increased performance, an HAC is used for data backup. The computer network in an HPC is mainly used for load distribution, while in an HAC data redundancy is created in a targeted manner. Thus, there is always a backup copy. So-called load balancers are used in both systems, which control the requests in such a way that they load the computers in the cluster evenly. In a network designed for data security (HAC), the hardware is relieved in this way, while in an HPC the read and write speed is also increased.

In practice, server groups are mainly used for sensitive IT services. Many large companies operate their own server farms in a network structure in order to be able to handle the large number of requests and at the same time to be prepared for failures. Alternatively, clusters can be rented: Internet service providers such as centron offer various server solutions, from simple servers to complex systems for customers with high data security and performance requirements.

You can find more information here.

Your tailor-made system: We configure your load-balanced, e-mail or failover cluster according to your wishes: Managed Cluster by centron.