The Wisdom of the Cloud

Forget blue-sky thinking; clouds are forecast for 2010 and beyond. Senior Editor Ben Thompson reports on why many believe the cloud will be the next big breakthrough in computing, and asks: why should your business care?

“Cloud computing is one of these concepts that come along every once in a while in IT and shakes up the foundation of the way we think about computing”
-Ric Telford, IBM

The sprawl of beige, box-like buildings on an unremarkable industrial park deep in Oregon's Columbia Gorge area hardly looks like a hotbed of technological revolution. The two huge units on the outskirts of The Dalles - each the size of a football-field - could easily be confused for a warehouse or assembly plant, while the low-rise admin block out front looks more like a construction project site office than the nerve center for one of the world's most innovative companies.

But look closer and telltale signs begin to emerge: the proliferation of power lines; the giant HVAC units sitting on the roof; the almost imperceptible hum of machinery as you approach. And then, of course, there's the security. Until recently, no-one in the area could tell you what was inside or even who owned the building, deterred by rigorously applied non-disclosure agreements and a strictly enforced no-access policy. ‘Keep Out' signs are plentiful - unlike company logos or other such corporate branding. But secrets don't stay that way for long in a small community; beyond the plethora of cameras, the formidable iron fence and the implacable security guards lies what the locals have come to call ‘Googleville' - and what many believe holds the key to the internet giant's phenomenal success.

Like the scene in Jurassic Park when you realize the resort is just a front for a highly developed industrial operation, visiting one of Google's many utilitarian server farms - the physical network that allows the world to search the internet, store files and send email, the sheer scale of which gives the company its competitive advantage - is a vastly different experience from stopping by the company's touchy-feely, beanbags-and-baristas headquarters building in Mountain View, California. For one thing, you've got more chance of gaining an audience with the President than you do of getting access to the endless rows of servers inside one of the company's data centers; Google guards its prize assets with all the attention to detail you would expect from a firm protecting its secret sauce. And while the Googleplex headquarters wins plaudits for its innovative working environment and army of boffins working tirelessly to develop new products and services, it is the data center that best represents the future of computing.

Why? Because the complex at The Dalles is just one of many facilities in far-flung locations, owned by some of the world's largest companies, that promise to deliver the next great shift in computing technology: the cloud.

Inside the cloud

As technology continues to evolve, more and more elements of our lives are being held not on our personal computers at home or in the office, but by websites on servers in remote locations we never even see. Data storage is cheap in these environments, often in the range of pennies per gigabyte - so cheap, in fact, that it is often provided in surprising quantities for free. Think the cloud doesn't affect you? Then think again. Those photos you uploaded to Flickr last night? In the cloud. Your Facebook profile? That's in the cloud, too. The YouTube video you posted of Uncle Brian falling over at cousin Susan's wedding? Sitting on a server somewhere in a facility very much like the one in The Dalles - again, in the cloud.

To the end user, the cloud is invisible. You access an application being run on a server operated by someone else, through the internet, as opposed to one stored on your own computer's hard drive. As far as Joe Public is concerned, the technology that supports the applications doesn't matter; it is the fact that the applications are always available that is important.

One of the earliest ways people used cloud computing was to store and access their email: rather than buying a server themselves, they signed up for an email account at a website like Yahoo or Hotmail, had their emails go directly to the provider's server and accessed them via the web. Gradually, that model came to be the accepted method of delivery for an increasingly varied array of services and products. For instance, people have quickly begun to take advantage of streaming video websites instead of downloading shows or recording them on DVR. Overall online video streams grew 48.9 percent from the middle of 2008 to October 2009, a figure that is still rising, and in 2009, TV on-demand provider Hulu saw total streams rise by a staggering 490 percent year-on-year. Meanwhile, streaming music sites such as Pandora and Last.fm have also seen growing interest as people choose to listen to music on-demand online rather than gather and maintain a large music collection on their hard drives.

"Cloud computing is one of these concepts that come along every once in a while in IT and shakes up the foundation of the way we think about computing," says Ric Telford, VP of Cloud Delivery Services at IBM. "It's a new style of IT delivery model that gives a highly scaleable, quickly provisionable, pay-as-you-go base to IT services - and does so in a way that allows customers to consume technology in a highly standardized way."

"Cloud computing is all about offering services over the network," agrees Telford's colleague Dennis Quan, IBM's Director of Autonomic Computer Development. "Service can mean anything from an email account, to access to computing facilities for doing a large-scale computation, to access to different information repositories and online video streams. No matter where you are, you can get access to your applications and your information, and using a laptop or a mobile device connect to the network. And these applications and collections of information are being managed for us centrally, somewhere up there in the cloud, so to speak."

A silver lining

The business benefits of such an approach are significant. Reducing - even eliminating - hardware and associated operating costs is clearly a no-brainer for corporate users, many of whom are already moving toward subscription-based software as a service (SaaS) models. Online business applications offered by companies such as Salesforce.com (for customer relationship management) and Workday (for human resources and financial software) not only replace expensive programs that would otherwise need to be run and managed by the company onsite, they can also reduce the need for corporate computer servers and the related costs of maintaining them. Instead, companies pay subscription fees for usage rather than licensing costly enterprise software. SaaS is a growth industry: a new study by Forrester Research concludes that even in the recent downturn, software-as-a-service providers were seeing double-digit growth in their subscription revenues.

Quan believes the variety of applications that could harness such cloud capabilities is substantial. "Online banking and e-commerce, web 2.0 social networking and collaboration, different types of real-time video and information exchange, as well as applications that support business, such as customer relationship management or financial modeling applications, could all benefit," suggests Quan. "We've also worked a lot with clients in the last year on different kinds of business use cases ranging from real-time collaboration to software development and test environments all the way up to different applications in financial services."

Amazon Chief Technology Officer Werner Vogels was recently named Information Week CIO of the Year for his role in bringing the online giant to the cusp of what he believes will be the computer industry's next major architecture. He believes companies large and small can benefit from the advantages on offer. "Cloud computing means highly reliable IT components can be used by startups, midsized and large enterprises alike to provide an environment in which developers or engineers no longer have to worry about managing physical resources," he explains. "Instead, they can use these resources as a service over the internet, and that will have a major effect on how applications are built. Applications will become more reliable, they will become more secure, they will become more cost-effective."

He feels such a development will free companies up to concentrate on the things that matter. "Just look at the amount of time that businesses currently waste on managing physical resources," he says. "Companies both large and small invest an enormous amount of manpower on just getting their infrastructure off the ground; factor in the cost of maintaining that infrastructure so that it is highly reliable and performing in such a way that customers can actually make use of it, and that's a big commitment."

Vogels likens the current situation to the dilemma faced by businesses at the beginning of the last century when, before the days of public utilities, most companies generated their own electricity. "Everybody felt that power generation was a core competency they needed to have because otherwise they couldn't perform basic tasks," he says. "However, as soon as public companies that provided electricity as a service came along, businesses couldn't wait to decommission their generators and focus on the things they did well. This is similar. Companies are investing a lot in managing a physical infrastructure when it doesn't actually help them build a better product for their customers."

Stormy weather ahead

Nonetheless, questions and concerns remain, not least around security, liability and interoperability - a point acknowledged by Telford. "How secure are public clouds? What if the cloud goes down? What if there's a disruption to the network? How do we solve that?" he asks. Security is certainly an issue: the findings of one recent survey from SaaS provider Mimecast show that security concerns were the leading reason given by respondents in all categories for not moving forward with cloud-based applications. Over 46 percent of respondents that had considered cloud-based applications chose security as the main reason for not moving forward, and this was true across a majority of industries, including financial services (76 percent), energy (75 percent), government (67 percent), retail (61 percent) and technology (40 percent).

"At the heart of cloud infrastructure is this idea of multi-tenancy and decoupling between specific hardware resources and applications," explains Datamonitor Senior Analyst Vuk Trifković. "In the jungle of multi-tenant data, you need to trust the cloud provider that your information will not be exposed." The answer lies in better security technologies around identity management, access control, perimeter defense and virus detection. "It's like in the days when people were uncomfortable giving their credit card details to e-commerce systems," says Telford. "We need to make sure people are comfortable leveraging cloud services over the internet too."

Rolf von Roessing, Vice President of not-for-profit IT governance association ISACA, believes business and IT managers should take the time to review their security systems and resources, and ask themselves whether their information security management can effectively defend their data and other assets in the cloud. And even if an initial budget to extend and enhance a firm's IT security resources to fully support the cloud is not apparently available, IT managers should not lose sight of the possibility that some of the direct cost savings that arise from cloud migration can be invested in security facilities. "It's important that in the rush to embrace the obvious benefits of the cloud, businesses do not overlook the security implications of extending their IT resource into what is, to all intents and purposes, a virtual environment," he says.

And security is not the only issue concerning IT experts. Vint Cerf, founding father of the internet and Google's so-called Chief Internet Evangelist, believes that while the cloud is already here, considerable work needs to be done in order for its true potential to be fully realized. "We don't have any inter-cloud standards," he warns, comparing the situation to the lack of communication and interoperability between computer networks he faced back in 1973 when designing the common protocols that revolutionized - and paved the way for - the worldwide web. "People are going to want to move data around, they're going to want to ask clouds to do things for them. They might even want to have multiple clouds interact with each other in order to take advantage of the computing power offered through such combinations. There's a whole raft of research work still to be done and protocols to be designed and standards to be adopted that will allow people to manage those assets."

However, not everyone is convinced that the industry is ready for standards yet. "Standards by definition are restrictive," says Trifković. "Consequently, people are questioning whether cloud computing can benefit from standardisation at this stage of market development. There is a slight reluctance on the part of cloud providers to create standards before the market landscape is fully formed."

A compelling case

Such confusion has done much to dampen enthusiasm around the cloud for many companies. Yet despite these concerns, there can be no doubt that the cloud is here to stay. Research just released by Resource on Demand, an IT recruitment firm, reports a record number of enquiries for Salesforce and software-as-a-service specialists during the first months of 2010. The financial and organizational benefits of switching to cloud-based storage and computing systems are clearly driving this surge in demand, and analyst firm Gartner recently predicted that within the next four years 20 percent of companies will have no appreciable local resources.

And like all great computing revolutions, it is the business case rather than the technology that is driving uptake. "With all due respect to the many wonderful providers out there, there's no breakthrough technology driving cloud computing," acknowledges Justin Ricketts, CIO at Susan G. Komen for the Cure, a not-for-profit organization at the forefront of the global fight against breast cancer that recently made the move to a public cloud solution. "For me, what's so compelling about cloud computing is the economics. If I can significantly lower my capex, and I don't have to manage it, don't have to maintain it, don't have to enhance it, and don't have to worry about new compliance coming down the pipe, then count me in. That's compelling. I believe the real paradigm shift in cloud computing is that computing boundaries are going to be defined by economics, not technical limitation."

Under Vogels, Amazon initially began investing in large scale cloud-based technologies to serve its own operational requirements, but following considerable internal success is now looking at using these to help other companies become more reliable, more cost-effective and more productive - without having the massive cost of building and maintaining those technologies themselves. "We can help companies become more reliable and more secure than they are now, at a much lower price point," he says. "Plus the pay-as-you-go model means that you pay only for those resources that you use. If you look at starting a typical new enterprise product, then often you have to get a large budget upfront to make sure you have the physical resources to execute on. By using web services, there is no upfront investment, and only if your product becomes successful will you carry the cost for the resources that you've been using."

He also believes the shift from a capital expense to an operational expense model lowers the risks for enterprises to do new product innovation. "There are no boundaries anymore for any company in any country to access these resources," he explains. "In some ways, cloud computing provides the democratization of business creation. You no longer need access to huge sums of money to get access to physical computing resources in order to get your business or product off the ground. I think technology development will shift - from actually having to manage physical resources and having multiple system administrators in order to keep your service going, towards building better applications for your customers. That's where the focus will be. Cloud computing will trigger a whole new range of application building that wasn't possible before because we were so focused on just getting the basics right. And we will see that applications become more available with better performance, because there will be more automation in terms of keeping these enterprise applications running - under all circumstances - in the cloud."

The future is now

These days, no computer user is an island. A recent study determined that 80 percent of the data used by business comes from outside the company, and cloud computing is the technical response to this reality. And while many don't see it as significantly different from pre-existing computing models such as hosted services and outsourcing, the point is that attributes such as usage-based pricing, shared resources, near real-time deployment and user simplification together make business and consumer cloud services easier and cheaper - and often better - to consume than through traditional delivery modes.

Paul Maritz is CEO of virtualization software firm VMware. He thinks the movement to the cloud is inevitable as more and more information - personal, professional and statistical - gets digitized. "Increasingly, individuals are characterized by a body of digital information," he explains. "And that information needs to live on over a period of decades - the rest of our lives - beyond the lifetime of any device you might have. Everybody is going to need somebody to be the custodian of their information. Just like we don't put our money under the mattress anymore, we put it in the bank, so most of us will become customers of an ‘information bank' - and in so doing become dependent on the cloud."

And herein lies the reason why cloud computing is so important: not because it creates greater efficiencies or scalability, or because the economics are so attractive; not because the vendors, marketers and media tell us so; but because it will ubiquitously come to touch every aspect of our increasingly digital lives. The pile of letters on the mat each morning has already given way to the e-mail database on the computer, which in turn is now making way for cloud-based Hotmail or Gmail accounts. The clipping left on the desk of a colleague becomes the attached PDF file, later to be superseded by a set of shared bookmarks, hosted offsite. The faded photos in an album are replaced by JPEGs on a hard drive, then a hosted sharing service like Flickr. The tilting CD tower - itself a digital replacement for our scratched and much-loved vinyl - gives way to the MP3-filled hard drive, which then inevitably yields to a service like Spotify: music that is always ‘there', waiting to be heard, somewhere in the cloud.

The point is, the cloud isn't just about the future; it's already here, present in both our homes and our places of work. Is cloud computing new? Not really. Will it reshape the way we use and buy IT power? Definitely. Like the internet in the 1990s, the cloud is an aggregation of ideas and solutions that just work. And the cloud will change the IT landscape like the internet changed the world. Adopting it is not a case of if, but when.