Server virtualization continues to revolutionize data centers around the world. No longer do system admins have to reserve an entire server for the company’s CRM, another for financial applications and yet another server for manufacturing or other mission-critical apps. Server virtualization in its three principal forms – full virtualization, para-virtualization and OS-level virtualization – all enable one physical server to act as several virtual servers, each running an array of application suites.
System administrators and IT management generally agree there are three key reasons to virtualize data center servers.
1. To optimize available resources
Modern servers with several multi-core processors and expansive disk storage offer far more processing power than most software application suites demand. Using that excess computing power and storage capacity by creating multiple virtual servers from a single physical server optimizes the available resources. Rather than running a rack of independent servers at, say, ten to 15 percent of their full capacity, virtualization allows one to use far more of the hardware’s computing power. While running a server at 100 percent of its capacity is not wise – because all the applications hosted on the physical server would slow down – it is not uncommon to fine tune the system to utilize up to 80 percent of its resources.
2. To reduce the cost of the data center
As applications are moved from the one-server-one-application model to virtual servers, the sheer number of physical servers can be reduced. The now-excess servers may well be used as backups to the load bearing servers that are virtualized. Or, excess servers may simply be retired, resulting in a reduced electricity and cooling bill each month. Gartner estimates that businesses save 7000kWh of electricity, or about $700 in annually for each physical server virtualized. Further, some businesses can see a 15:1 reduction in servers once they are virtualized. This means system Administrators can manage virtual servers more effectively, leading to reduced maintenance costs and more strategic initiatives from the IT department.
3. To improve business continuity
Keeping mission-critical applications up and running is the most important role of the data center. A failed server or application lights up the tech support lines almost immediately and can cost an enterprise dearly unless the applications are brought back online in short order. Virtualization software provides fail-over or “migration” features that almost immediately transfer processing from a disabled physical server to its redundant backup server, resulting in almost no noticeable delay or interruption in most cases. If you are already running VMware vSphere you should consider upgrading to vSphere 5 and taking advantage of VMware Site Recovery Manager 5 for a comprehensive Disaster Recovery and business Continuity Solution.
Picture by *hokkey