Thumbnail: 

2009 was the Year of the Cloud for most tech vendors. But if you ask 27 different vendors to define "cloud 2computing," you're likely to end up with 27 different answers.

By its strictest definition, anything "in the cloud" uses computer resources not physically present at the point of origin. A mainframe accessed by "dumb-terminals" (a typical setup in past decades) might qualify using that definition.

But that doesn't help us understand what's going on in the early years of our new century. Cloud computing in 2009 is the greatest shift in the computing paradigm since individual firms started aggregating computing power in the form of mainframes and servers and allocating it to their employees. The "cloud" analogy works because much computing nowadays is conducted wirelessly, and the main avenue is the formless internet. When college students made up something called "Hotmail" in the 1990s, it was so you could communicate by email outside of an educational or corporate network. Maybe this wasn't strictly "cloud computing," but it reflected the concept.

Cloud evolution

Nowadays, many of us use our email archives as a primary record-keeping mechanism, and our historical email files are an important resource.

But what happens if the email files are not backed up regularly? Whether your primary email is a part of a corporate network or simply your personal copy, odds are good that you have your email set to delete the messages from the server as soon as they are downloaded to the PC. And even if a copy of the emails may still exist somewhere in the bowels of the IT department, recovering these emails can be a major issue.

But if you're using a network-based service, such as Gmail, then all of the email would be "safe." This has the advantage of potentially recovering not only the correspondence itself, but also the vast majority of important files.

Even though accounts from providers like Gmail don't have licensing fees, they're a way of storing emails and attached files in "the cloud".

On the enterprise-level, cloud computing can be used for far greater things. The next year or two may see radical shifts as firms like Google and salesforce.com gear up their cloud-centric products, as netbooks continue to proliferate.

It's an interesting time for technology, and some of Asia's tech experts are intrigued.

"At the beginning of 2009, I thought cloud computing was just hype or at best that it was simply another name for outsourcing," said Linda Hui from F5 Networks. "There were not many companies in the market actually providing cloud computing services. The name we heard frequently was salesforce.com which I feel - and many others feel - is 'software-as-a-service', a fraction of what cloud computing is about."

"However, as 2009 progressed and more and more companies - such as IBM, AT&T, Amazon and SingTel - began to provide cloud computing services, the situation changed. There was the realization that a temporary infrastructure could be set up and that potential customers could pay for what they use, and that virtualizing resources was becoming a reality. So now I see cloud computing as a more concrete proposition and I believe that it will really take off in 2010."

From servers to service

Even Microsoft has gotten on the bandwagon. "Cloud computing is about taking the complexity out of IT without sacrificing the capabilities, said CEO Steve Ballmer in an interview. "At Microsoft, this means that anything that has been a server needs to be a service so that customers can choose whether to run on-premise, in the cloud or take a hybrid approach."

Ballmer added that "around 70% of customers using Microsoft Online Services have moved from legacy platforms: for example, Proctor & Gamble, Glaxo Smith Kline, Coca-Cola Enterprises and Blockbuster Video have all made the move."

According to a recent IDC research sponsored by Microsoft, looking into the adoption of SaaS, 26% of Hong Kong companies with 25-500 employees are already deploying or considering deploying SaaS. "That's three times the regional average," said David Hooper, information worker group lead, Microsoft Hong Kong. "The vast majority of the cloud applications are communication and collaboration based such as email, calendaring, web conferencing and instant messaging."

Which projects are right for the cloud? Some IT functions are perfect for the cloud, while others need to stay in your data center.

Budget-minded CEOs are telling IT managers to look into cloud computing to reduce the amount of expensive hardware running their data centers, CFOs are interested because they've heard the model can slash costs associated with new IT projects, tech-savvy employees are asking for it because they think it sounds cool. IT departments large and small feel obligated to at least look into cloud computing's potential to save money, reduce overhead and increase efficiency and flexibility.

What's more, those IT shops that drag their feet might find overeager users are beating them to the cloud, warns James Staten, an analyst at US-based Forrester Research. For example, "application developers are using the cloud and not telling IT," he said. To avoid being caught unaware, IT should take the lead in deciding what goes into the cloud and determining how to get it there, said Staten.

Emerging best practices

In a report published in September, Forrester Research outlined the following best practices for cloud computing:

• Conduct functional and scalability testing and development work.
• Deploy short-lived and highly volatile Web applications.
• Run quick, grid-type high-performance computing analysis.

But where to start? What's the best way for an IT manager to determine whether his company's corporate culture is suited for shipping computing tasks to web-based third parties? What expectations should service providers be required to meet? How should the success - or failure - of a cloud computing project be measured?

These are not questions to be taken lightly, since the success or failure of a company's foray into the cloud will influence corporate perceptions of the model going forward.

The Corporate Executive Board, a research and membership organization designed to support the functions surrounding CEOs, has studied corporate adoption of cloud computing through its Infrastructure Executive Council and its Data Center Operations Council, both of which are headed by practice manager Mark Tonsetic.

Tonsetic's advice to IT managers: find a project that supports a business opportunity and could be easily moved into the cloud to cut costs and resources - something that doesn't involve core competencies and moving it offsite shouldn't create a security risk. In other words, find a project where moving some or all functions to the cloud would improve the bottom line but the company wouldn't face disaster if security or availability was compromised.

Tonsetic isn't alone in advising companies to tread lightly into the cloud - security risks are created when companies move sensitive information beyond the limits of their own data centers. And, as proven by a number of recent high-profile outages of cloud services provided by Google, Microsoft and others, availability is a real concern.

"Look at your portfolio of applications and services and decide which are commodities, not core competencies," Forrester's Staten advises. "Those are your candidates for cloud."

Worth doing right

Embarking on a cloud computing project may take extensive research and preparation, but the payoff can be significant when everything is done correctly. In order to realize the promised reductions in cost, companies need to make sure they pick the right projects to send to the cloud.

"From a business point of view," said Tonsetic, "it's important to understand cloud computing's capabilities and match those to opportunities, then evaluate different technologies and vendors."

NEXT: China's SaaS market pegged at $170m in 2010 [1]