Collaboration

Services consumers and developers must now mount pressure for cloud computing neutrality

Dana Gardner explains why he think it's important to foster an open market and encourage early adopters -- especially developers and independent cloud services providers -- to vote mindfully with their participation (and dollars) among the burgeoning cloud entities on the Internet.

 

This is a guest post from Dana Gardner of TechRepublic's sister site ZDNet. You can follow Dana on his ZDNet blog BriefingsDirect, or subscribe to the RSS feed.

Sure, most people instantly get the need for network neutrality, but what about cloud neutrality?

Just like we'd be loathe to tolerate any one (often the only available) Internet provider from qualitatively managing our traffic and packets use based on their singular business objectives, we should also be concerned about any cloud provider exerting too much influence or setting de facto standards early on that diminish the cloud services market as a whole.

Now, the Obama Administration has enough on its plate, so I'm not advocating any regulatory or commerce enforcement policies to define or advocate cloud neutrality. But I do think it's important to foster an open market and encourage early adopters -- especially developers and independent cloud services providers -- to vote mindfully with their participation (and dollars) to establish and nurture broad openness and interoperability practices among the burgeoning cloud entities on the Internet.

If an open Internet has been good for sustained productivity and innovation, which few refute, why wouldn't cloud services also benefit from an open market environment -- at least through a formative stage (or two)? Wouldn't what's good for the popularity of the pipes also be good for promoting the widest consumption of the water?

Let's still favor the advance of general productivity on the Internet over more narrow commercial interests, even as we enter the cloud services phase of the Internet, eh?

Shouldn't a network infrastructure often described as "public" -- hence the common icon of the Internet as a puffy cloud -- become the substrate for an intensely fertile marketplace, and just not a handy subsidy for any number of, albeit competing, roach motels? The best form of competition comes from the hotels competing but also with low barriers of entry and exit over a long period of time. Choice is essential, not just among vendors but on how those vendors behave as a group.

The cloud services marketplace is not just a new Monopoly board in the sky, it's still the product of the World Wide Web. If you have to go through the cloud to reach the services, then the services themselves are a product of the cloud, and not the other way around.

Things in the nascent cloud services ecology are moving rapidly, so now's the time to set the proper perspective on what works best for the buyers and users of cloud services, as well as the commercial interests setting up shop along the Information Superhighway. Remember that metaphor for the Internet? I think we should think of a cloud superhighway in the same way. Now it's more than information, but it's just as or more important to the public good. There's a public interest in seeing this succeed for the highway travelers, which include big businesses, as well as those few building the toll booths.

There are some dramatic recent developments that point to how rapidly cloud things are shaping up:

  • IBM's Lotus brand is bringing a lot of what we know as Notes/Domino services, a longtime enterprise groupware leader, to cloud-based delivery. Think of it as a big nest of .nsfs in the ether (and that's data up there, folks!).
  • Engine Yard's Vertebra has extended cloud neutrality into its Ruby and Rails development and deployment solutions. Write anywhere, run anywhere, change anywhere, integrate anywhere ... repeat.
  • Sun Microsystems buys Q-Layer. Let's hope that Sun gets cloud "open" from the start this time, unlike the 12-year Java will be open someday saga (and keep those license fees coming).
  • Zimory, a Deutsche Telekom spinoff, unveils Zimory Public Cloud beta, an online marketplace that brings together buyers and sellers of cloud-based computing resources.

There have been warnings about a potential and troubling lack of choice in cloud options, notably from Richard Stallman. And there have been major movements by vendors not known for their allegiance to openness first and profits later, including Apple and Microsoft, into the cloud model.

So even though things are moving fast and at the most impactful levels of the global IT business, there's very little being said and done about preserving the neutrality of the Internet economy for the cloud economy. And I know it's hard to actually define neutrality. But like pornography, I know it when I see it.

Better yet, I know non-neutrality when I see it. We should all be on the lookout for non-neutrality in the cloud ecologies, and seek and reward alternatives. Blog about these distinctions. Look to the decades-old Internet example for guidance. It really worked and keeps working.

That does not mean in any way outlawing good old fashioned capitalism in the cloud ecosystem. It means making savvy choices that favor data portability, and recognizing that APIs that carry over from one hosting provider to another make for good market drivers that entice more consumers that can exercise more choice. The pie needs to grow first, and the market leaders can seek domination in some way later when the playing filed is established and perhaps somewhat level.

Enterprises and small to medium size business especially should advance their long term interests as they examine and adopt cloud-based services to make sure they are not trading short-term savings in a recession for long-term IT lock-in. Once you're in the roach motel, you can't get out. And they can raise the rent (maintenance fees) to just below your cost of exercising painful choice for a long time. You may be familiar with this IT supplier dynamic.

There is a better path, and we've seen it with the Web: A modest, market-driven level of mutually beneficial interoperability of services and applications, data portability in its deepest forms, SLAs that clearly spell out the terms of engagement and what is acceptable in terms of services and data ownership.

These cloud terms of engagement will be tough and complex. We're in some uncharted territory here. Can you own a business process even if the cloud provider owns the constituent services? Yes, I believe you can, and should. Get it in writing, though.

So more than any regulations or broad policy dictates on the best practices for cloud computing, we need good licenses and a clear and understood framework for cloud ecology best practices that protects the users and developers, as well as the providers. The goal is to make strong enticements for all the participants in the ecology, not just a few or in a grossly inequitable way. We'll need escape clauses, too, just in case.

Indeed, the value and probity of cloud use licenses must be weighed against the IT total cost equation, including the cost of switching and the costs of integration. That is, if I get cloud services cheap, how much will that cost me in the long run? And is this and does this become a better deal than the traditional on-premises, per processor or application licensing models?

In short, we need the ability to calculate the cost-benefit analysis of modern IT that includes the new cloud computing options. And therefore we need to know the true costs of cloud computing -- including how open it really is -- to proceed. The more open, the less risk, and so the more overall adoption based on an understood cost-benefit projection.

Let's look at cloud services as hugely promising, perhaps the best alternative for IT resources and support for a number of applications types and certain business use cases. But let's not get lulled into treating a cloud provider relationship any different from any other business deal. Let's get the terms down, and vote well as consumers. It's in the best interest of the vendors, too, they just can't do this without us. Literally.

Let's leverage the fact that the Internet has set a powerful and essential precedent that upholds and protects an online market's open development as fundamentally more important than any one company's ability to stake out a claim and horde all the gold dust. Open markets are the best way to allow the miners, prospectors, shovel sellers, and real estate interests to all grow and prosper. And openness will allow the cloud market to reach its full potential fast, through unfettered innovation from all quarters.

Like with the Web and Internet over the past 15 years, the power of choice and unfettered innovation and dynamism of sufficiently neutral cloud markets should be the best guide of how the cloud future shakes out productively. In this economy we really need a new and huge productivity boost from IT lest we all get pulled into the downward spiral.

5 comments
WebWatcher
WebWatcher

Dust Cloud, cloud of smoke, head in the cloud, every cloud has a silver lining, clouded judgement,thunder clouds stratus clouds, cirrus clouds, cumulus clouds, cumulonimbus clouds... All acceptable uses/derivatives of "cloud". "Cloud computing?" Pure marketing tripe and draws immediate derision on those who parrot it blindly.

Tony Hopkinson
Tony Hopkinson

Service providers will determine the market, by providing those services that come up with the best profit. They'll go for ecomomies of scale, merge and dominate, until there is no choice but them. Then all that economic benefit to consumers will dispappear in the cloud of smoke from burnt currency. Who are you trying to kid?

htmapes
htmapes

You need to read fewer press releases. Cloud computing, at this stage, is the equivalent of the word "portal" during the latter years of the dot-com boom. It's a marketing placeholder, nothing more. It is no more realistic to expect portability between cloud environments than it would be to expect an application compiled on a Mac to run on Wintel or a C## app to be an easy port to Java. What cloud computing will do is abstract geography and physical hardware into availability and function. The end result will be to increase utilization and make the economics more granular -- you pay for what you use. Every vendor will produce this in a different way based upon their existing capabilities and target market. Amazon is going after a no-real-SLA solution based on their existing platform. My companies has designed its cloud product for the enterprise that needs predictable performance and regulatory compliance. Companies will build internal environments that will be required to support heterogeneous, legacy environments alongside virtual server farms. In short, there is a radical difference between transport and computing. Standards will emerge, but they will be based on market penetration, de facto dominance and regulatory compliance.

Justin James
Justin James

This made little sense to me. "Cloud neutrality"? Heck, I challenge you to find database neutrality, or even file system neutrality! The idea that something as complex as cloud computing is (or can be) can simply be "neutral" is quite baffling. First question... what standards would he like these clouds to adhere to? Last I checked, there are no standard CRM interchange formats, so I guess by definition Salesforce.com can't be "cloud neutral". Ditto for just anout anything else other than email, FTP, and a few other well defined applications that are based around protocols. Or, to put it another way... If, after 50 years of not having "neutrality" amongst various applications, what makes him think that it is possible to have it happen in the "cloud"? J.Ja

Editor's Picks