"Data centers are not immune to the menace of hidden costs. Fortunately, many of the extra expenditures don’t result from the baffling rabbit hole otherwise known as fine print. A good amount of them stem from the realities of the data center environment, and often from developments that make data center outsourcing and managed services a better overall investment. However, unplanned costs can hit businesses at the wrong time, as they often seem to do, and cause problems for short-term productivity and long-term profitability. Avoiding hidden costs is mostly a matter of education – the higher one’s awareness, the easier it is to understand what factors can result in extra costs. It will also make it easier to avoid them.
1) Primary Storage Bottlenecks
Bottlenecks are one of the top causes of unplanned downtime and system slowness, which can lead to decreased productivity. One virtualization vendor estimated that more than 70 percent of application performance support desk calls pertain to issues with data storage and access. Virtualized servers have somewhat helped alleviate the problem and are one of the chief reasons why third-party data centers can be more reliable than on-premises data storage. However, virtual server environments can develop their own problems. A common issue results from the deployment of too many virtual machines per data store, TechTarget reported. I/O bottlenecks created by demanding virtual machines can be difficult to detect, often resulting in different degrees of crippled virtual applications. Complementing existing hard drives with solid-state drives can help alleviate system latency.
2) Water Consumption
The problem of scant water resources is becoming a global issue, which could make targets of companies, including data centers, which use large amounts of water in their facilities. The new National Security Agency data center will reportedly use 1.7 million gallons of water per day, to cite one high profile example. Many communities are mulling higher water connection fees for data centers, which would lead to higher costs of operation for these facilities. There’s movement in the industry to substantially reduce water consumption – data center providers like Google, RagingWire and DuPont Fabros have installed water recycling programs on campuses to use less water. However, the life aquatic should still be on the radar for both owners and tenants.
3) Redundant Staff
This isn’t a pro-robots argument, per se, but extra IT staff covering data centers are growing increasingly redundant. The traditional on-premises data center model demonstrates a clear demand for IT personnel to oversee the system. Many of the skills needed to develop a strong support system for an in-house data center are superfluous in a managed services environment. Organizations have two options to trim redundant staff from their payroll. The first is obviously to dismiss these employees or replace them with cheaper outsourced labor, acknowledging that much of the work is actually performed within the data center and requires less oversight than was initially believed necessary. The second is to build on employees’ existing skill sets with redefined roles. Big data is one example in which IT skills are valuable building blocks in the effort to transform information into business decisions.
4) The Wrong Place
One of the extra costs of data centers could be hiding in plain sight. Because location can be a hidden extra cost factor, it’s difficult to determine how much the choice of region is affecting the bottom line. Conventional wisdom supports close proximity of data centers to company offices, but this notion can end up costing the business dearly. There are some more radical options out there – Iceland is justone of the alternative locations promising low cooling and disaster recovery costs. However, many of the top markets in the U.S., like Ashburn, New York/New Jersey and the Silicon Valley, offer network connectivity options that can further defray costs, while the Midwest and mountain regions provide space and stable climates."
