Though cloud computing often is little understood, it is among the hot technologies for 2009, and it stirs up strong emotions among both supporters and detractors. For example, during a meeting with analysts in September 2008, Oracle CEO Larry Ellison ranted, "We've redefined cloud computing to include everything that we already do. I can't think of anything that isn't cloud computing with all of these announcements. The computer industry is the only industry that is more fashion-driven than women's fashion. Maybe I'm an idiot, but I have no idea what anyone is talking about. What is it? It's complete gibberish. It's insane. When is this idiocy going to stop?"
Yet later that same month, Oracle announced that several of its products would run on Amazon's Elastic Compute Cloud, EC2. Further, Oracle has assured the public that it has more cloud products under development, and it has a well-stocked Cloud Computing Center on its Web site, complete with product datasheets, podcasts and white papers. Reluctantly or not, Oracle has jumped on the cloud bandwagon as much as anyone.
Here's why cloud computing — loosely defined as Internet-based development and use of computer technology — has a future and the likely reason why Ellison doesn't like it: Cloud-based services so far are cheaper than traditional IT products, such as large in-house databases.
Cloud computing users leverage servers, applications and/or storage hosted by a provider (or by their own company, in the case of internal clouds) and are billed strictly for what they use each month, like a metered utility. So if a company is launching a new product that requires a large database, it doesn't need to go out and buy one; it can use one hosted by a cloud provider such as Amazon, Google or Vertica. In contrast to large up-front costs, initial costs are small because data lookups are few. If the new product turns out to be a success and lots of customers access the data, costs will go up, but then revenues should grow as well.
Still Ellison isn't the only one questioning the wisdom of computing in the clouds. Also in September Richard Stallman, founder of the Free Software Foundation and creator of the computer operating system GNU, said that cloud computing was a trap aimed at forcing more people to buy into locked, proprietary systems that would cost them more over time. "It's stupidity," he told U.K. newspaper The Guardian. "It's worse than stupidity — it's a marketing hype campaign."
Stallman is a privacy advocate who believes that when people use Web applications they give up control. That ship, however, has already sailed — most companies and people utilize Web applications and have grown used to the idea of working over the Internet.
In fact in early December Claus Mortensen, IDC's principal for emerging technologies advisory services, predicted that IT cloud services will form 25 percent of all incremental global IT spending growth by 2012. In addition Merrill Lynch analysts said that same month that by 2011 the cloud computing market will reach $160 billion, including $95 billion in business and productivity applications. Merrill uses IBM's Blue Cloud servers to build and evaluate new risk analysis programs. Morgan Stanley uses Salesforce's Force.com cloud offering for recruiting applications.
Nasdaq Looks to the Clouds
When Claude Courbois, associate VP, product development, Nasdaq Data Products, created a new product called Market Replay that enables brokerage firms to show customers and regulators that best-execution requirements were met for a given trade (view the tool at https://data.nasdaq.com/mr.aspx), he looked into buying a database to support the product. "Naturally database solutions are the first things people look at, because they're a safe solution," Courbois notes. The price for a dedicated database, however, was prohibitive, forcing him to put the project on hold for several months. Then Courbois realized that if the process work could all be handled on the client side, Nasdaq could use a simple data storage product such as Amazon Web Services' S3 storage cloud.
Today Nasdaq stores many terabytes of Nasdaq, NYSE and Amex data in Amazon's storage cloud; according to Courbois, Nasdaq adds 30 gigabytes to 80 gigabytes of data every day to the cloud, about 300,000 flat files each representing 10 minutes' worth of trading activity on a stock. The data retrieval time, he reports, is less than one second, and the system scales instantly.
Market Replay reconstructs the environment around a trade by pulling all the historical market data related to that trade and creating a screenshot of market conditions at the time of execution. This can be sent to regulators or customers who question a trade. "Often customers or regulators will call about a trade that happened several months ago," Courbois explains. "The fact that we're able to keep so much data online indefinitely means the brokers can quickly answer a question without having to pull data out of old tapes and CD backups."
More important, Nasdaq never pays for one byte more than it uses, Courbois says. "If we built this ourselves or used a standard ASP [application service provider], we'd have to ask for more space than we initially need and pay for all these empty terabytes until we fill them up," he relates.
On the other hand, the initial monthly bills Courbois received from Amazon Web Services were as low as $5. "I never had to buy $20,000 worth of hardware or enter into a big contract," he says. "Even though we're in a big company, every new project is a start-up, and you want to avoid situations where you have to plunk down a bunch of money to move forward." And if the new product wasn't a success for some reason, Courbois could simply delete the data and cancel the Amazon service.
To purists, Nasdaq isn't technically using cloud computing (although it plans to soon) — rather it is using cloud storage. The difference is that Nasdaq didn't build its Market Replay application in the cloud, nor does the user interface exist in Amazon Web Services. The application was developed using Adobe Flex for the Adobe Air runtime, which provides more computing power than a standard browser, Courbois reports. But, he adds, Nasdaq plans to develop future applications in Amazon's Elastic Compute Cloud. For instance, the exchange will create an application to let users search historical market data and perform calculations on it, according to Courbois.
For now, though, Nasdaq's next cloud project is taking Market Replay to Europe, using the same Amazon S3 cloud to store European stock data, Courbois says, adding that Nasdaq plans to have the product up and running in the first quarter of 2009.
Nagging Cloud Doubts
Cloud doubters often cite a lack of control as a hurdle to cloud computing: What if there's a problem? Who's going to fix it, and when? Courbois says that although Nasdaq hasn't had any IT issues with respect to Market Replay, Amazon has responded to questions quickly and has an excellent support Web site. "It's not a hand-holding situation — they don't advertise it as such, and you can't expect it to be," Courbois stresses. "It's not like paying for a data center or server-side ASP."
Other cloud skeptics worry about security. But Courbois claims this is a bit of a red herring. "There are ways of securing data and then you can put it anywhere," he says, pointing out that if Nasdaq were to put proprietary data on S3, it would encrypt it first.
Forrester Research analyst James Staten, who wrote in a recent report that enterprise IT is not ready for the clouds largely because of security concerns, says, "It's not that [the cloud providers] can't handle security, but they need to articulate it better." Staten notes that he has seen firms build and manage Monte Carlo simulations, risk analysis scenarios and forecasting models with cloud computing.
Nonetheless Nasdaq's Courbois concedes that there are limits to a cloud's usefulness. For instance, Nasdaq has no plans to run its stock market in the cloud in the foreseeable future — "That has to be in our own environment," he says.