NewsFactor Network | The Great IT Complexity Challenge

The Great IT Complexity Challenge
By Joe "Zonker" Brockmeier

April 30, 2003 4:00AM

Digg It!

Standards will be critical in driving adoption of autonomic computing. "One of the big issues is how this is going to evolve," said Gartner analyst Tom Bittman. "We will end up with a standard, but it will be after a bloody battle. This will be messy."

Technology is supposed to help simplify transactions and increase the speed of doing business, but often that is not the way it works. While technology certainly can speed things up, it also can impede progress. A company can become so tightly bound to any given technology that it loses its agility. Change then becomes a difficult, slow march.

As for the idea that technology reduces complexity, nothing could be further from the truth. Integrating all the various and sundry systems that are supposed to simplify business operations is a complex task in and of itself. Then there is the need for ongoing maintenance and periodic modifications to adapt the systems to current business requirements -- which tend to change more rapidly than the systems that support them.

Major IT leaders, including IBM, HP and Sun Microsystems, are stepping up to the plate, developing autonomic-computing systems that are designed to simplify the management -- and ratchet up the responsiveness -- of enterprise technology solutions.

The Promise

What is autonomic computing? In a nutshell, the term means "the ability for IT to become self-managing," according to IBM's program director of autonomic computing, Miles Barel. Of course, the definition is somewhat changeable from vendor to vendor -- in fact, so is the name.

IBM has coined "autonomic computing"; HP prefers "adaptive infrastructure "; Gartner calls it "real-time infrastructure"; and Sun's solutions are grouped under the "N1" umbrella. No matter what you call it, the goal of this new technology is the same: to make enterprise systems easier to manage, more reliable, more flexible and more efficient with less human intervention.

Essential Elements

There are four elements for a system to be considered "autonomic," Barel told NewsFactor. It should be self-configuring, self-optimizing, self-healing and self-protecting. There are also different levels of autonomic computing, Barel noted, adding that its full potential is far from fully realized.

One way of considering autonomic computing in terms of maturity. "We start with 'basic' -- not autonomic at all -- where the complexity is most notable," Barel explained. "Second is 'managed,' where you use a common view of data coming from all components and manage from a central location -- but [this is] still done by humans." These levels are already widely available.

The next two levels are less common, according to Barel. "The third level is called 'predicted.' The technology has the ability to take data from the systems and environment and analyze that data and make recommendations to IT staff to effectively run that environment." (continued...)

1 | 2 | 3 | Next Page >