Cloud Computing

They say that soon we will do our computing by renting memory and processing on a vast cloud of computers out in the network. They say this is an idea whose time has come. Of course ”they” have said ”the time has come” about other ideas, and some of those things didn’t happen then, or sometimes ever. But this time it’s for sure. They say.

I remember the old days of time-shared computing. Big, expensive computers were kept behind big glass walls, tended by shrouded acolytes. For the rest of us, it was ”keep your hands off.” You rented computation by the second for your dumb terminal.

The miracle of Moore’s Law changed all that. Computing equipment has become so cheap, the junk piles are full of it. Who would have ever thought of a terabyte of data just lying around the house? So why should we go back to renting computational power, this time from a network cloud?

Apparently it’s precisely because of Moore’s Law. Processors and memory are cheap these days--and really cheap when you buy, run, and maintain them by the zillions. And the clouds at companies like Google and Amazon are living things, constantly being replaced and upgraded with the latest cheap stuff.

The advantages for us, the cloud’s clients, include the ability to add more capacity for peak demand, to flexibly experiment with new services, and to remove unneeded capacity when demand slackens. Within the giant cloud, the laws of probability give the service provider great leverage through statistical multiplexing of all these varying loads. The cloud is also easier to manage--you can install a single software patch to cover all of a company’s users, for example.

Those seem like compelling arguments, so what could be wrong with this picture? For one thing, despite the existence of encryption and access-control technologies, many organizations seem reluctant to put their precious, proprietary data out in a public cloud. There is a widely held impression that this might not be a good thing to be doing. Also, there is a practical and administrative problem--once all your petabytes of data are out there in the cloud, can you ever get them back?

While corporate IT organizations ponder the advantages and drawbacks, as an individual user I have a different problem--giving up my autonomy and control. I’ll admit from the start that this is a psychological hang-up; nevertheless, it’s there. I rather like having my own cycles and bytes; I like being able to tend my own computing garden. Basically, the question is: Do I trust these guys?

Over the years, I’ve always been both a system designer and a system user, and I’ve seen that the two sides are not always congruent. The designer feels that the best system is a closed one, where all the features and interfaces are under his control. I don’t think that designers have evil intentions or even that they’re always trying to make more money for their companies; often the urge to control is simply a part of the quest for engineering perfection.

No, it’s not the designers I distrust so much as the businesspeople. I’ve seen the cellphone that takes pictures that can be sent only to my computer, over the contracted carrier, at a certain cost per picture; the built-in GPS that requires a subscription service; the e-book reader whose crisp electronic page can display only books that come through a particular vendor. All of these may be perfectly good systems, but they don’t give me the flexibility I want. So why should I trust the cloud-computing service providers?

There is a counterargument, of course. The cloud advocates say that the open market will create open systems; that openness will give some service providers a competitive advantage. That may already be happening, as there are different offerings that put the user at both the bottom of the stack, where the access is effectively to a bare generic computer, and at the top, where the access is to customized applications.

So is cloud computing an idea whose time has come? It’s your call. In the end, that’s the way it works.