Cluster hosting

Cluster hosting involves the use of several servers to avoid the problem that can arise when a single server is used to host websites and results in high availability hosting.

When you have a single server then a hardware failure can cause all websites hosted on the server to spontaneously go offline, a situation which is intollerable for many hosting clients. Clustered hosting allows traffic to be moved to a different server if one goes offline, or stops functioning properly.

Grid web hosting often uses banks of Linux/Unix web servers running Cpanel and WHM, with a bunch of RAID drives. High availability clusters (HA clusters) of servers are sometimes known as failover clusters and utiles nodes or redundant servers.

In a high availability cluster, based around a heartbeat network connection, redundancy eliminates single points of failure and are vital for international ecommerce websites, corporate sites, and government websites that need to deliver pages under extreme visitor traffic circumstances at all times.

What are the options when you decide to buy clustered server hosting?

The number of nodes in a high availability website hosting server cluster varies, but two is very common, in order that an appropriate level of redundancy is provided. Active and passive nodes are common, with N+M, N+1 and Nto NO active/active a good idea. Some applications need to be redesigned in order to work in a clustered setting and you need to think about the implications for software licenses for products such as anti virus, firewall and Microsoft SQL Server.

Redundant switching fabric and firewall proxy technology is employed in the intelligent routing that is commonly found in clustered hosting network layers. The real advantage of clustered hosting is that when multiple websites share such an environment even the largest user of resources just uses a fraction of the amount available, reducing the risk of spikes caused by unexpectedly successful websites, denial of service attacks, hacking and network attacks.

In website server networks and server farms load balancing is the system used to spread work between two or more servers, cpus and hard drives, when you need to host high traffic websites, big Internet Relay Chat (IRC) networks, DNS Servers, high bandwidth File Transfer Protocol (FTP) websites and NNTP servers. Resources are shared, response time is reduced and reliability and redundancy are enhanced. A multilayer switch is usually used for the purpose of load balancing, with the load balancer sending page and file requests to backend servers.

Other factors to take into account when load balancing among clustered high availability hosting servers include TCP offload, Content aware switching, Asymmetric load, Priority activation, HTTP caching, Content Filtering, TCP buffering, Priority queuing, Client authentication, Spam filtering, Firewall, Distributed Denial of Service (DDoS) attack protection, Programmatic traffic manipulation, SSL Offload and Acceleration, HTTP compression, and HTTP security.