Don’t Virtualize Waste

Server virtualization has been proclaimed as a significant way for organizations to increase operational efficiency, cut IT costs and reduce their environmental footprint. it has been said that the effective use of virtualization technologies can reduce server energy consumption by up to 82 percent . unfortunately, many companies that have embraced virtualization have yet to save what they expected. why? They continue to virtualize waste.

For server virtualization to reach its full savings potential, virtual servers must be carefully managed and used productively. Companies must visualize useful work across IT assets to achieve the full efficiencies from virtualization projects and get the ROI they expect.

Virtual Sprawl: Virtualization should Save, Not Cost IT

though virtualization often delivers on its promise of reducing data center power consumption and generating initial energy savings early in a project, the operational costs associated with running a server are many times that of the energy cost. it is now so easy to rapidly provision a new virtual server that an exorbitant number of virtual servers are created – many more so than physical servers would have been without virtualization. IT can spawn a new virtual server to test out an idea or try an application before committing to roll it out and, unless tight controls are maintained, these virtual servers can be left behind unused when the testing has finished. Although they don’t take up physical space or require extra cooling, they tie up valuable disk, memory and CPU resources that could be more usefully deployed for productive work. Plus, they are simply waste.

This is known as virtual server sprawl, and it is starting to become a real concern for organizations around the globe.

Unforeseen Costs of Virtualization

Much attention has been paid to the energy and space savings a company can achieve through virtualization. This has created a misperception that running virtualized servers is inexpensive when there are actually many hidden costs.

One such cost is related to facilities, server hardware and storage hardware. Organizations consume more energy when they purchase more high-specification host servers to support the growing number of virtual machines (VMs). The standard specification of a server purchased to host virtual machines is usually much higher than a standalone x86 server. Hosting multiple virtualized servers requires more CPUs and significantly more memory, both of which contribute to power requirements and heat output.

If an IT department virtualized 100 percent of its physical servers today, it would actually be operating more servers than it started with. Despite the fact that the physical and environmental footprint have changed, the IT department still must monitor, patch, secure, backup and license both the virtualized servers and now the host servers.

as the number of VMs increase so does the required storage, resulting in a possible increase in storage hardware. Backing up these additional images takes time, and this can sometimes be difficult to achieve in the given maintenance window.

Yet another hidden cost is related to software licenses. Organizations cannot assume that a license from a physical server can be converted to a VM license without additional costs. some software vendors require a license for each server on the virtual host  – leading to spiraling licensing costs.

on top of this, virtual sprawl inevitably leads to an environment running applications not in use. even if license cost is no longer a factor, those installations still need to be properly managed or the “in use” count of licenses will be inaccurate. Waste can also be represented in terms of host resource (power, shared disk, RAM) and IT effort. If virtual machines continue to run after they are no longer needed, the initial savings gained from virtualization will be lost. Gartner calls this paradox the ”Virtualization Catch 22”.

Virtualization also intensified the complexity of server total-cost-of-ownership (TCO) models. Internal chargeback mechanisms that worked when each application was conveniently ring-fenced by the hardware on which it ran are obscured now that one physical server hosts multiple applications belonging to different business units. to compensate for this, some organizations have adapted their models by simply applying the cost of operating a physical server to each virtualized server, an imprecise approach to the issue.

participate in immigration news polls