||Add To My Personal Library
December 22, 2006
Vol.28 Issue 51|
Page(s) 9 in print issue
The Biggest IT Issues
Experts Weigh In On Today’s Most Pressing Demands
Like isolating a snowflake in a blizzard, identifying crucial issues in the data center can prove immensely difficult—especially when every issue appears more critical than the next. But the inability to determine the IT issues deserving of the most attention can spell big trouble for IT managers down the road.
Whether its power, heat, security, or nearly any other issue, there is no single area that consistently proves to be the biggest IT issue across all data centers at any given time, primarily due to the wide range of configurations found in the SME space. However, experts have little trouble identifying certain standout issues that materialize in todays data centers more regularly than others.
According to Andrew Hillier, co-founder and CTO of CiRBA (www.cirba.com), a developer of data center intelligence and consolidation software, one of the top issues facing managers today is server sprawl, which he says is actually the root cause of more problems than people realize. By their very nature, sprawling environments in turn spawn sprawling costs, and organizations are beginning to focus on reducing those costs, particularly when it comes to software licensing rationalization and heat and power costs.
But after looking deeper, it also becomes clear that reliability and service levels are also highly linked to the complexity and diversity in an environment, and there are big wins here, as well, Hillier says. People are getting tired of having too many links in the chain and having to constantly reverse-engineer their data center every time something goes wrong.
Server sprawl, says Hillier, is a natural result of open systems. Because oversizing a server is a prudent approach to building responsiveness that can handle an applications early life cycle, the one-box-per-app mentality has been common practice for years now. Smart enterprises can effectively absorb any excess capacity in that box as the application matures and managers increasingly understand its workloads. Enterprises that cant perform those actions face problems.
We view this as an intelligence gap, and because organizations have not had visibility into their excess capacity and analysis that allows them to safely make use of it, the result has been a proliferation of underutilized systems, Hillier says.
The ever-increasing need for security is always high on the priority lists of IT managers, but according to Roger Thompson, CTO of Exploit Prevention Labs (www.explabs.com), a developer of safe surfing software and services, browser-based exploits are a particularly crucial issue.
The problem is that firewalls and perimeter defenses are great for keeping network worms and hackers out, but anything browser-based goes straight through the firewall because its initiated from a trusted place—the inside, Thompson says. You simply cant patch 1,000 PCs every month when Microsoft does its monthly update, and the bad guys understand that. They know that even a simple exploit will be effective in corporations long after Microsoft has released its patch.
Thompson says that most people still think that exploits lurk in shady areas, such as adult Web sites, but most corporations will admit that employees pick up infections when innocently surfing the Web—or surfing nonadult sites. This occurs due to hackers who break into poorly defended Web sites and add code that connects back to an exploit server.
No longer is it a constant challenge to find space for power and cooling—instead, its a challenge to find enough power and cooling to accommodate the equipment that fills that space. Ahmar Abbas, the head of ManageInfra (www.manageinfra.com), the remote infrastructure management division of IT services company SlashSupport, says that the introduction of blade servers and similar technologies has dramatically increased server density.
Abbas explains that the number of power-consuming elements inhabiting data centers has also increased—a group that includes firewalls, switches, load balancers, intrusion detection systems, DNS appliances, and mail appliances. Further, the growth of raw storage requirements is also increasing in enterprises of all sizes, with IT managers deploying larger storage and backup systems and constantly upgrading capacity.
The over-capacity that existed after the dot-com bubble has been exhausted, and we are fully in the supply-constrained market, Abbas says. Consider that Google, Amazon, and others are building large data centers in Oregon—right next to hydroelectric plants. Congress is even holding hearings and may in the future mandate certain efficiency requirements for data center equipment. So, at least for the next two years, we will remain in this market.
Bill Mansfield, solution architect for Logicalis (www.logicalis.com), a provider of high-performance IT integration solutions, predicts that the most important upcoming IT-related issue for data centers is the storage management challenge of IT service management.
Were moving past the era of monolithic big-box storage into a new age where storage virtualization in the SAN commoditizes the big boxes, enabling and encouraging their acquisition from the lowest bidder, Mansfield says. Storage administrators [SAs] will no longer routinely manage their storage arrays, so when something does need to be done at that level, the SA will lack the skill to reliably do the work.
Further, administrators will need to develop skills to handle new virtualization appliances, while management will face pressure to maintain or reduce staff for storage management. Although storage management applications do exist, most of them provide only device management and monitoring centralization and lack planning and workflow support, Mansfield says.
The result will be an environment that is more complex and less understood, prone to unforeseen adverse interactions among its components, he says.
by Christian Perry
Additional Issues |
Heres a look at other expert opinions on todays biggest IT issues.
One item that is getting more and more attention is the amount of energy that is used in data centers—not just for the computers systems but the supporting infrastructure, as well. As the cost of energy continues to rise, CFOs and CEOs are looking at their big power bills and asking, Who is using all this electricity? Right at the top of the list is the data center. Often, the word comes down to cut your power usage—a demand that is easy to make and hard to do.
—Robert Rosen, immediate past president of IBM user group SHARE and CIO for the National Institute of Arthritis and Musculoskeletal and Skin Diseases
One of the most important issues facing IT managers today is gaining control of their IT infrastructure. With the declining cost of hardware, the number of servers and storage in the SME data center continues to grow at a rapid pace. With this growth, data centers in SMEs are starting to look more and more like the data centers larger enterprises had just a few short years ago. With this growth comes both the urge and the necessity to manage the infrastructure so that visibility into the network is kept high; information is kept secure and viable; and problems are fixed before they snowball or are prevented entirely.
—Tony Barbagallo, vice president of product management and marketing, GroundWork Open Source (www.groundworkopensource.com)
IT managers face simultaneously internal and external issues that will pull on their resources. . . . The most significant internal issue is centered about the customers (users) and their expectation. More and more, this expectation is expressed in terms of how IT connects its services to the business side of the house. It is a strategic partnership, and it is about impacting the bottom line. Externally, it is about security and privacy issues. We want the community [to be] confident in our ability to protect their information and value their privacy.
—George Kahkedjian, CIO, Eastern Connecticut State University