I’m a firm believer that if one wants to increase the odds of being wrong, all one need do is make a few predictions. Now that I’ve gotten that out of the way, let me just say that I think that this could be the year for utility computing. Sure, last year we saw some offerings in that space but not really of great overall impact in the industry. After all, we’ve been talking about it for more than five years. Could it be that the way is now clear?
Utility Computing is more than the nexus of SOA, Data Center Virtualization and Grid Computing but, in practical terms, it is no less. For big hardware and software players, the way may be clearer, but it is still littered with legacy systems. Last year, I heralded Amazon’s Utility Compute offerings as serious utility computing for the “masses.” Could it be that a grass roots move toward utility compute will be the spark that triggers widespread adoption? My guess is yes, where new applications are concerned but only maybe when it comes to enterprise wide legacy systems.
It isn’t just Amazon that is expanding into utility, eBay, Google and Yahoo have long been headed in that direction as well, just not as generically. While Amazon creates generic offerings as a utility compute provider, eBay refines its SOA while working with third party vendors. EBay is busily refining the automation and interface model it uses to interact with its customers. This clearly serves to streamline its operation. So, for the biggest players on the Web, utility computing isn’t just an architectural shift driven by the desire of a few techies to experiment with new technologies, it’s a better way of doing business.
So what about the big enterprise legacy systems? Modernizing most of these systems will be expensive and tedious but ultimately necessary since many have grown brittle with age. As utility computing proliferates and proves its worth on the Web, the need for transformation will become even more compelling. Ultimately the popular shift to utility computing may force enterprises to modernize.
So with respect to new Web 2.0 applications, startup companies and assorted new business applications, we will see a strong showing for utility computing and utility computing ready solutions this year. For big enterprise legacy systems, this may not be the year for utility compute, but I think it may well be the decade.