Evaluate Weigh the pros and cons of technologies, products and projects you are considering.

How big data and mobility fuel the need for open cloud architectures

Open cloud architectures let providers scale storage and integrate with enterprise storage platforms -- two big requirements in the big data age.

The ability to analyze and interpret data can open up countless growth opportunities for service providers, university research labs, enterprises, startups and small businesses. Those opportunities, however, aren't without challenges.

Managing terabytes, petabytes and exabytes of data will soon become the norm, and cloud providers need to think about how they use and permit access to that data. Developing those strategies will be essential for any cloud provider looking to tap into big data analytics for internal use or looking to build customer-facing services on top of that data.

Whenever you start handling increased volumes of data, how they're stored almost always raises questions about logistics. Who can access the data and how much you expect it to move around both become major issues. Open cloud architectures offer a possible solution by providing the ability to scale storage more efficiently and integrate with enterprise storage platforms -- two major requirements in the age of big data.

As a use case, mobility is a prime example of how the cloud can solve big data challenges. For starters, the limited amount of storage on mobile devices means tablet and smartphone apps often need to "borrow" additional storage and computing power from the cloud.

There's another area -- addressing a bigger issue -- where an open cloud can play a role: interoperability. Although user behavior is similar across mobile devices, the data structures and streams collecting the relevant data about this behavior differs greatly. Moreover, each mobile analytics service provider has its own method for storing user behavior data.

Additionally, users are interacting with the various native applications on their mobile devices, and they require a degree of interoperability to move data between different platforms -- an issue that grows more urgent in a world of multiple, competing mobile app platforms and ecosystems. Mobile users expect to access any publicly available content or apps, regardless of the mobile platform. Users of one wireless carrier want the same access to applications that they would receive from another provider, and they want to buy music from an app store from their Windows desktop computer and then sync playlists to a mobile device or music player running Android or iOS.

Open cloud bridges proprietary mobile platforms

This need for platform agnosticism in a data-driven world is fueling an open cloud movement that not only gives users the rights to move data as they wish, but also enables federation, provides a complete view into the infrastructure environment, and creates a collaborative developer community. To fall back on the previous examples, mobile apps need to work with services consistently -- irrespective of the mobile operating systems they run on or the cloud provider behind them. There is, and will continue to be, a constant tension between cloud providers that may attempt to lock in customers through proprietary platforms and the push from the marketplace to maintain ubiquitous access.

The term open cloud -- used often but rarely defined -- refers to a broader set of architecture characteristics, features and capabilities. An open cloud has open application programming interfaces. An open cloud gives users the ability to move data as they wish. An open cloud is federated, enabling an organization to move the same data between and among various cloud environments. An open cloud has full view into the infrastructure environment. Finally, an open cloud has a collaborative developer community -- not one based on hosted, legacy software.

As enterprises face economic pressures, resource limitations and skill shortages, interest in cloud computing -- particularly the open cloud -- is increasing. And as the debate between private and public clouds pushes enterprises toward hybrid cloud adoption -- a presumably more flexible cloud computing model -- the fear of vendor lock-in persists. An open cloud, however, provides the interoperability to eliminate these concerns.

For example, Africa is one of the key emerging markets for cloud services. Meanwhile, public cloud market penetration currently is at 70% in India, with private cloud penetration at 24% and hybrid cloud at 6%. This demonstrates how more organizations are accepting the need to take advantage of open connectivity.

Building a community to produce a ubiquitous cloud computing platform can make a simple, scalable and feature-rich service available to any developer, retailer, service provider, enterprise or small business that can run it.

About the author:
Dennis Quan is vice president of SmartCloud Enterprise at IBM. He and his team built IBM's first cloud in 2007 as part of a collaboration with Google and the National Science Foundation. He has held numerous technical leadership positions in IBM over the past 14 years, including most recently as lab director of the IBM Tivoli China Development Laboratory in Beijing, Shanghai and Taipei, where he worked extensively with clients in the Greater China region across various industries on cloud computing and integrated service management projects.

Dennis holds both a Ph.D. in computer science and a bachelor's degree in mathematics and chemistry from the Massachusetts Institute of Technology.

This was last published in December 2013

Dig Deeper on Telecommunication networking

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

This is an excellent insight Don !!! THANK YOU !!!
Dear Don, this is very insightful. Being an IBMer myself, am proud in knowing what IBM is doing to improve this area. I did some courses at MIT and your article shows the condensed version of what is goign on in there !!! but i still feel miles to go before we can say we are at the forefront...we must detach machines/computers out of this growing data itsretention and analysis.. because machines limit the capacity to measure the data..data is being created by somany things in this world...and unfortunately,currently we are not able to even measure/analyse the data being created by the small machines we created..and to understand the behaviour of nature/animals and humans( which are totaly intervowen) the data available to our machines( generated by our own machines) is not sufficient. so I guess instead of limiting the capcity by adding a mchine elements, i suggest we use non machie methods( which wer followed and implemented and got results- Example: pi value, space distances and calendars--- in india in ancient times) toi decipher an dexcel in analytics !!! only people like you who have dwelved a lot into data sceinces can get these things done !!! Kindly have a look in that direction too..!!! THANK YOU once again for an insightful article !!!