The great OpenStack balancing act

Can a vendor consortium cooperatively develop a cloud operating system without going off track? The new OpenStack Foundation and a small army of developers aim to find out

The OpenStack project got its own open source Foundation last week. This so-called cloud operating system, launched under an Apache 2 license by cloud provider Rackspace and NASA two years ago, will be governed by an independent organization that bears some similarities to the Linux Foundation.

In fact, Alan Clark, who was just elected chairman of the board of directors of the OpenStack Foundation, also serves on the board of the Linux Foundation and helps oversee the Suse Linux project. And there's another familial resemblance: Just as few would download the Linux kernel from and put it into production, few enterprises are likely to download and deploy the raw OpenStack bits from

[ Stay on top of the current state of the cloud with InfoWorld's special report, "Cloud computing in 2012." Download it today! | Also check out our "Private Cloud Deep Dive," our "Cloud Security Deep Dive," our "Cloud Storage Deep Dive," and our "Cloud Services Deep Dive." ]

Instead, vendors will package OpenStack and incorporate it into various products, much as various vendors develop Linux distributions or embed Linux in data center hardware. CanonicalRackspaceRed Hat, and Piston Cloud already offer OpenStack downloads. And public cloud providers such as HP are launching their own offerings based on OpenStack.

But despite this proliferation, OpenStack remains at a very early stage. There will undoubtedly be exceptions, but among enterprise customers, significant adoption of private cloud OpenStack solutions for production purposes is probably a year or so away.

Risks and rewards
That timeframe points to a stark difference between OpenStack and Linux: The latter had a decade to mature before it was enterprise-worthy. OpenStack is taking shape before our eyes, mainly through code contributions from established vendors. The stakes are very high, because OpenStack provides a management framework for nearly every virtualized resource in the data center.

To me, this breadth -- and the ripple effects of decisions made in code -- make OpenStack a unique experiment with its own special set of risks and rewards.

The main OpenStack reward for customers is indirect: Collaborative development of the core OpenStack bits should shorten the time-to-market for private and public cloud solutions, the adoption of which should increase IT efficiency and reduce costs overall.

OpenStack proponents also say that sharing the OpenStack bits will prevent cloud lock-in, but that benefit may be limited. Those who have been around this tree before (think J2EE application servers) understand the inherent contradiction: Each OpenStack cloud software or service provider will differentiate with special features that, if you use them, will inhibit portability.

1 2 Page 1