Not so long ago, users depended on server farms – featuring applications deployed across multiple servers – to protect their systems against server failure. The thinking was that if all applications were divided among numerous servers, and one server failed, the other servers would continue to operate and daily business would continue with minimal (hopefully) inconvenience to the organization.
But over time server farms grew so large that they required a huge investment in physical resources and management time. Virtualization successfully addresses these issues by featuring all the necessary multiple server images – called virtual servers – hosted on a single physical server. This effective solution greatly reduces the time required to manage physical servers and significantly reduces company investment in physical hardware.
The key to the virtualization concept is the SAN – or Storage Area Network. The SAN hosts any and all data that the company needs. In addition, the SAN is scalable, so it can expand in capacity as the company’s requirements expand.
In order to compensate for the fact that the dependency on multiple servers has been eliminated, a great deal of redundancy is built into the new architecture. Duplicate switches, duplicate servers, and a SAN are set up with multiple connections to each other in order to effectively provide failover to a stable platform in the event of a failure of any single device or connection. A typical array of equipment with a high level of redundancy follows: