The company’s forthcoming line of “Blue Cloud” products is in some ways easier create a business case around than it would have been a few years ago. There are companies, like Google and Microsoft, which are poster children for how well this kind of thing can work. Soon there will be others, probably appearing at IBM’s next Information on Demand conference, which will offer their own testimonials. After watching the pendulum swing back and forth between mainframes and client-server, IBM admits it is trying to do with cloud computing what it did with Linux: announce to enterprise IT managers that it’s officially smart to at least start considering this approach.
Most likely Blue Cloud’s initial push will happen through IBM’s Global Services customer base, which presumably already trust the vendor and might be predisposed to the kind of flexibility it affords. That it is tailoring specific software and hardware to assist with deployment – and basing them on open source Xen and PowerVM – is a lot more interesting than simply forming a partnership with VMWare.
IBM, of course, is not the first company to put forth this vision of cloud computing, it’s simply talking the most loudly about it. Dell already offers a cloud computing package, and Sun Microsystems’ Jonathan Schwartz raised the concept several years ago. At the time, however, he pointed out issues around security and pricing – tricky matters when your architecture is so highly distributed – that might still stymie enterprise deployments. (IBM has yet to release pricing.)
If companies don’t show interest in cloud computing it may be because they don’t face the kind of compute workload crisis that the likes of a major search engine or social networking site do. In most industries, proper forecasting can give you the kind of heads-up you may need to allocate resources effectively, whether it’s the U.S. Thanksgiving weekend in the retail sector of RRSP deadline day in finance. For these kinds of businesses, retooling their architecture around a compute cloud may be something they consider only once they’ve experienced serious downtime or an unprecedented stain on their network resources.
It’s possible that in due course, many enterprises will set up their networks much like a Google or a Facebook, which predominantly Web-based applications and staff deployed more remotely than locally. Cloud computing may make sense then, most a lot of enterprises are less than half-way there today. (Finding out they need to re-think their data centre strategy could, in fact, be a deterrent to Web 2.0 among businesses.) As for the potential data centre cooling costs cloud computing might provide, I suspect companies will likely opt for more energy-efficient hardware before they take this kind of step.
When you stare at real clouds as a child you can imagine them to look like all kinds of people, places or things. The obstacle for IBM and other vendors entering this space is to help customers see cloud computing as only one thing: a future-proofing architecture they can roll out with minimal disruption.