What do Dell, Red Hat, SuSE, Cloudscaling, Piston Cloud, Cisco and StackOPS all have in common? They each have their very own OpenStack distribution and this is potentially a major problem for the nascent OpenStack development community.
Based on the attendance at last month’s Folsom Design Summit, it is clear that OpenStack has the potential to be a real game changer in the Infrastructure as a Service (IaaS) industry. The OpenStack Foundation has been favorably compared with its Linux and Apache predecessors as “an advanced example of open source governance” according to Jonathan Bryce, the new Foundation’s executive director. All that is well and good, but there seems to be a growing number of companies that are not sticking with the main (or the so-called trunk) OpenStack program and making the decision to fork off their own distributions for sale to their customers. Soon, if OpenStack follows the same development path as Linux, there will be hundreds of distributions with subtly different features only noticeable to hardcore Linux kernel engineers, and just plain confusing to the majority of the user community.
Since there is no real dominant market leader yet and the code base is still relatively unstable, each vendor is jumping in and trying to build a stable enterprise-ready OpenStack platform. While there are valid reasons for taking this approach, each new distribution has a net negative effect on the overall community because it can be seen as a distraction from the real work that needs to be done to stabilize and consolidate the underlying code base. At the Summit during the Analysts View of OpenStack panel, Krishnan (Krish) Subramanian argued that there was a perception in the enterprise that the major OpenStack vendors are running the show and that transparency is important. On the technology side, which distribution becomes the de facto enterprise standard is important because these customers are used to the Red Hat subscription based support model. This translates to a requirement for an OpenStack distribution to be transparent to the user, stable and as close to the main development trunk as possible; a tall order indeed given the current state of the code base.
For the cognoscenti, who appreciate the subtle differences between the distributions, more power to you, but for the rest of us, this smacks more of marketing hype, a way to create vendor lock-in where it doesn’t belong, an attempt to stabilize the code base, or some combination of all three. At this early stage in the development cycle, such obvious attempts to grab market share do not seem to fit into the true spirit of the Open Source community.
About the Author
Beth Cohen is a senior cloud architect for Cloud Technology Partners.
What do you think?