It’s easy to find good things being said in the IT media about virtualization. It’s less common to encounter some sobering thoughts. That’s why a particular dire litany in a recent Eaton Corporation white paper — written by Eaton’s manager of distributed power solutions, Chris Loeffler — caught our eye.
“For all its advantages, virtualization brings some unique challenges,” he writes. “Overall data center power consumption will be lower, but each server will draw more power. There will be fewer servers, but each one will be more critical than ever. Applications can be dynamically moved around as needed, but the support infrastructure cannot do the same. Data center footprint will be smaller, but overall data center efficiency might still be suboptimal.”
We don’t want to imply that virtualization isn’t a great technology. But Loeffler’s comments are reminiscent of the scene in the movie “Pee Wee’s Big Adventure” when Pee Wee announces, “It seems everybody has a big ‘but.’”
People who don’t understand computers often see them as tools of voodoo, capable of magical — and sometimes cursed — behavior. Longtime IT veteran Jaime Henriquez, who holds a doctorate in technology and culture, recently compiled a list of common computer “superstitions” for TechRepublic.com.
“These are the users who have memorized the formula for getting the computer to do what they want but have no clue how it works,” he explains. “As in magic, as long as you get the incantation exactly right, the result ‘just happens.’ The unforgiving nature of computer commands tends to feed this belief.”
For example, refusing to reboot is a major superstition. “Some users seem to regard a computer that’s up and running and doing what they want as a sort of miracle, achieved against all odds, and unlikely ever to be repeated — certainly not by them,” he writes. “Reboot? Not on your life!”
Another example is believing the computer has a personality. “This is the user who claims in all honesty, ‘The computer hates me,’ and will give you a long list of experiences supporting their conclusion,” Henriquez says.
Or believing the computer is all-knowing. “Things this user says betray the belief that behind all the hardware and software there is a single Giant Brain that sees all and knows all — or should,” he writes. “They’re surprised when things they’ve done don’t seem to ‘stick,’ as in ‘I changed my email address; why does it keep using my old one?’”
“Once on the path to magical thinking,” he continues, “some users give up trying to understand the computer as a tool to work with and instead treat it like some powerful but incomprehensible entity that must be negotiated with. For them, the computer works in mysterious ways, and superstitions begin to have more to do with what the computer is than how they use it.”
I ran across a couple of articles about innovative Data Center cooling and power strategies based on good old seawater and tidal waves.
It seems the island nation of Mauritius (yes, you can find it on Google Maps) plans to develop a system to use seawater air conditioning (SWAC) to support data center tenants. The concept taps deep-water currents that bring colder water within two miles of Mauritius.
In Scotland, huge data centers will share energy generated by a tidal power generation project. This article also references Google’s proposed floating data barges that would use oceanic power.
Unfortunately, we aren’t close enough to the coast to harness sea power in Beaverton, but certainly our mild climate offers some power-saving cooling alternatives. (Stay tuned…)
Many IT-related decisions ideally hinge — or should, anyway — on the total cost of ownership (TCO) of a company’s data center.
Easier said than done.
“There are no recognized standards for measuring the TCO of the physical infrastructure of data centers,” says a report from American Power Conversion (APC), a Rhode Island-based global leader in network-critical physical infrastructure solutions. “Simple methods of summing various capital and operating expense items do provide insight into total cash outlay, but they do not account for the utilization of the equipment.”
“Consider the case of two data centers, each 100kW capacity and built identically,” the report states. “In one case the data center is fully utilized to 100% of the space and power capability; and in the other case the data center has only a single rack with 2kW of equipment. While the cash cost of operating these two facilities over their lifetime is comparable, the useful return on that investment is totally different.
“In the totally utilized case, the TCO of the data center is spread over a large amount of equipment providing useful services. In the lightly utilized case, the entire burden of the expense of the physical data center infrastructure must be borne by the single rack.
“When TCO of physical data center or network room infrastructure is measured from the point of view of the useful work performed, namely how much IT equipment is supported, underutilization can drive extraordinary cost,” according to the report.
The APC report concludes that the single largest TCO cost driver is the unabsorbed overhead cost of underutilized IT infrastructure.
For companies considering the cost advantages, say, of an internal IT infrastructure versus an outsourced one, it’s important to also remember our recent posts on this blog on “Data Center Overbuilding,” pointing out that the average data center is oversized by three times its design value, and at commissioning, tends to be oversized — read “underutilized” — by up to 10 times.
EasyStreet and other managed services providers (MSPs) across the land are on a mission to explain the important differences between an MSP and a typical IT “outsourcing” arrangement. Boston-area consultants THINKstrategies, Inc. recently discussed some important differences in a white paper.
The paper quotes Gartner research on why at least half of today’s traditional IT outsourcing agreements fail to meet their objects:
- Traditional outsourcing typically focuses on reducing customer capital investment in IT and ongoing IT expenditures, but fails to proactively mitigate risks and optimize performance.
- Traditional outsourcing often includes strict contractual terms and conditions restricting the range of issues an outsourcer will address without added fees.
- Organizations become overly dependent on the outsourcer because they have taken ownership of their IT assets and displaced their internal IT staff.
- Traditional outsourcing arrangements generally last a minimum of five years and include significant penalties if they are terminated or restructured.
“Contracting with a managed services service is a more flexible form of out-tasking,” according to THINKstrategies consultants. “Customers can use MSPs to provide varying levels of support from tier-one issue identification to tier-three problem resolution and incident management.
“The MSP typically augments rather than fully replaces the in-house IT function. However, in some cases, SMBs decide to have the MSP assume full responsibility for their IT operations,” the paper continues. “MSPs also provide a more flexible portfolio of services which can be acquired incrementally. The contract structure of managed service agreements is also more flexible, and a standard contract period is typically less than five years.
“Managed services also enable SMBs to convert their IT capital expenditures into operating expenses and to establish a more predictable month-to-month cost structure for their IT operations, without the day-to-day operational hassles of managing their own IT,” according to THINKstrategies.
To download a copy of the THINKstrategies white paper, click here.
