RECIPEAs time goes by, the relative cost of available CPU, RAM, and storage capacity continues to drop. As a result, many PC and server users have more computing horsepower than they can fully utilize. For system builders, this presents an opportunity to help customers make more complete use of this idle hardware. One solution that can help is virtualization technology.
By consolidating servers with virtualization, system builders can help their customers deploy applications more efficiently, using less hardware. System builders can set up a Windows server and a Linux server on the same physical machine, for instance, and allocate resources between them as necessary. In this way, virtualized servers can save customers a lot of money, and system builders can take credit for streamlining their customers' IT infrastructure. This should win more business opportunities in the long run. It should also reduce potential service calls by putting less hardware into the field to run the same applications.
Microsoft, among others, offers a product that addresses this need: Virtual Server 2005 R2 Standard Edition. In this Recipe, I'll discuss the ins and outs of server virtualization, and explain how to install Microsoft's Virtual Server 2005 R2 Standard Edition software to create a virtual server.
The proliferation of 64-bit computing hardware and software means system builders can now install extremely high-performance systems. For example, today's servers can address more than 4 GB of RAM. CPU clock speeds now approach 4 GHz, with multiple cores on each chip able to tackle additional computing demands. Storage capacity has skyrocketed, too: Terabyte storage arrays are now within the reach of system builders who serve the small and medium business (SMB) market. Until quite recently, these performance levels were limited to only the largest of enterprises. But no more.
Another phenomenon is contributing to the under-utilization of available hardware, too: Today's Commercial Off-The-Shelf (COTS) software is typically compiled for maximum compatibility, instead of for performance and resource utilization. In fact, these applications are typically optimized for use on a single machine with one or two CPUs and up to about 2 GB of RAM. But if your systems have more power than that, they're probably not being fully utilized by the software.