||Add To My Personal Library
December 12, 2008
Vol.30 Issue 50
Forecasting The Cloud
A Look At How Cloud Computing Will Impact The Enterprise Of The Future
AS ENTERPRISES continue to hop on board the ever-expanding phenomenon of cloud computing, it is becoming difficult to imagine a future without the cloud. But seasoned executives who have endured failed computing fads in the past take cloud computing with a grain of salt. Will the cloud persist? If it does, how will it impact the enterprise?
• Both cloud and in-house applications will continue to exist in future enterprises.
• Although any application could exist in the cloud, not all applications will make sense in that architecture.
• Smaller enterprises are more likely to utilize the cloud in the near future.
“SaaS [software as a service] and cloud computing will enable enterprises to outsource more non-core IT functions to third-party providers and to reduce IT infrastructure costs,” says Tyler Newton, research director and general partner with Catalyst Investors (www.catalystinvestors.com), a private equity firm focused on growth sectors of the economy. “It will also increase the availability of cutting-edge technology to businesses of all sizes, reducing the ability to use technology for competitive advantage.”
Cloud On The Horizon
Today, the cloud allows enterprises to enjoy extensive IT services without the hassle and worry of managing or supporting the infrastructure around them. Experts predict that in the future, those benefits will continue, but traditional software architecture won’t be heading for the exit any time soon. Newton says enterprise software will come in three forms: SaaS purchased from third-party providers, internally written software hosted on a cloud computing platform, and traditional on-premise enterprise software. He adds that standardized software platforms increasingly will be purchased on a SaaS basis, while heavily customized software will be a mix of SaaS, cloud, and on-premise.
George Symons, CEO of Yosemite Technologies (408/737-3311; www.yosemitetech.com), agrees, explaining that enterprises will have a mixture of in-house applications and applications delivered as a service. Those applications that are standalone—from a data perspective—and designed for a single purpose or department are good candidates to be delivered as a service, he says, because they will allow IT to be flexible and reactive without requiring a large staff. Other complex, mission-critical applications will require data sharing and dependency among applications.
“I don’t think [the move to the cloud] is going to be exactly the same as in the early industrial age, when manufacturers produced their own electricity and then slowly converted to public power grids,” says Jeffrey Pattison, CIO of INTTRA (www.inttra.com). “Some enterprises will put their core applications in a cloud, but there will be a big concern for reliability. Just as we have backup generators for electricity today, we need a backup for cloud-based processing.”
Although Newton predicts the cloud could limit the use of technology for competitive advantage, Symons sees opportunity for an advantage with those complex applications that are intertwined. “This is where the enterprise might be able to have a differentiable advantage over competitors or simply require disparate applications to work off of the same data,” he says. “These applications will continue to be installed and managed in the data center. I expect there will continue to be issues around corporate governance and compliance that will [require] many enterprises to house these applications and their data in-house.”
According to Patti Dock, executive vice president and chief marketing officer of DataMotion (www.datamotion.com), people talked about cloud computing as a delivery model just a few years ago, but now there’s a huge emphasis on the cloud being a cost-containment model. “I predict that within five years, we will see cloud computing being used as a productivity model. In this same time frame, we will move from discussing products in the cloud to productivity in the cloud, and the emphasis will shift from availability and scalability to transparency and from moving data through the cloud to a world where intelligent information knows how to traverse through the cloud,” Dock says.
At the cloud’s current rate, it’s not outrageous to imagine that nearly all enterprise applications could find a home there. However, that’s not likely to happen, experts say. Joe Sturonas, chief technology officer for PKWARE (www.pkware.com), predicts that the cloud won’t contain all applications, partially because of intellectual property and data security concerns.
“Another complication we are seeing today is that some cloud providers are starting to evaporate—not just getting bought or participating in mergers, but disappearing altogether. [Some providers are] being reasonable and giving users plenty of time to remove their data . . . but those businesses will need to find an alternative service now,” Sturonas says.
Applications that require high availability and guaranteed response time are unlikely to find their way to the cloud, Sturonas says, because the cloud model uses too many shared and unreliable resources, such as the Internet itself, which he says can become congested, and servers, which are hosted and include other applications or customers that can starve applications from CPU, I/O, and memory resources.
DataMotion’s Dock says that while smart engineers can devise methods for moving any applications to the cloud, there will be practical reasons for enterprises to decide not to move everything there, such as security concerns, convenience, speed, and the general notion of practicality. “Think about having someone having access to your mainframe application from the cloud. I doubt operations people will be excited about the prospect. So even if it is possible, it is unlikely that all applications and services will [be] moved to the cloud,” she says.
Smaller Heading Skyward
At some point in the future, cloud computing is bound to demand the rapt attention of nearly every enterprise. But moving forward, Catalyst Investors’ Newton predicts that cloud computing will prove most compelling to smaller businesses with limited resources to staff and IT budget.
“Larger enterprises will be the last to fully embrace cloud computing and SaaS due to the IT investments that they have already made,” Newton says. “Cloud computing will, however, dramatically reduce the cost of R&D for developing new software applications, which should spur innovation.”
by Christian Perry
Headed To The Cloud? |
In theory, almost any application could exist in the cloud, though observers might question the value of its presence there. Predicting the future of the cloud can be difficult without basing that prediction on any real metrics, so Yosemite Technologies CEO George Symons (www.yosemitetech.com) looks at an application’s cloud viability from a more practical standpoint: data ingestion.
“If data is created incrementally and accessed incrementally, then the cloud is perfect. If an application ingests and/or requires access to large amounts of data bursts and those bursts are consistent, the application is not well-suited for the cloud,” he says.
For example, if a terabyte of data must be moved from system to system for analysis, the cloud won’t be an ideal route for the data’s travel, even with the increases we can expect in bandwidth over the next 10 years, he says.