Network & ServersClustered hosting technology is designed to eliminate the problems inherent with typical shared hosting infrastructures. This technology provides customers with a “clustered” handling of security, load balancing, and necessary website resources.
A clustered hosting platform is data-driven, which means that no human interaction is needed to provision a new account to the platform.
Clustered hosting "virtualizes" the resources beyond the limits of one physical server, and as a result, a website is not limited to one server. They share the processing power of many servers and their applications are distributed in real-time. This means that they can purchase as much computing power as they want from a virtually inexhaustible source, since even the largest customer never consumes more than a fraction of a percent of the total server pool. Customer account changes (to add new resources or change settings) are propagated immediately to every server in the cluster. This is different from typical shared hosting architectures that usually require changes to a configuration file that becomes live after the server is rebooted during off hours, or are pushed on a cyclic basis every few hours.
Multiple tiers of security are integrated into the clustered hosting platform. In a typical hosting environment, the security layer is usually not integrated in the platform. The stock solutions used for shared hosting do not solve core issues around integrating security between the application and the operating system. At best, most typical hosts will implement a firewall solution, and weaknesses inherent with the operating system will remain exploitable to those that penetrate the firewall.
Clustered hosting network layer protections employ intelligent routing, redundant switching fabric and built in firewall and proxy technology. Clustered hosting provides considerable advantages over traditional hosting architectures in mitigating denial-of-service attacks and other network attacks because such attacks can be dispersed over a large pool of servers, and if individual hardware components are impacted by such attacks, they automatically fall out of traffic handling during the attack.