OpenStack Foundation Launched

The OpenStack Project (previous CloudAve coverage), so far under the “control” of Rackspace, announced their intention to launch a foundation which will control the governance and trademarks of the project. This is a big news and warrants a short analysis because I have talked about this issue when it erupted after Rick Clark’s departure and when they bought Anso Labs. The announcement will be made tomorrow at the OpenStack Conference in Boston and the foundation will be formally launched in 2012.

This marks a dramatic shift in Rackspace’s approach to the idea of an OpenStack Foundation. When calls were made by prominent open source pundits like Simon Wardley and Sam Ramji for an independent foundation, Rackspace pushed back hard against the idea touting the Apache license as a potential weapon against any hijack attempt by Rackspace. In fact, after the last OpenStack Design Summit and Conference I was convinced that OpenStack needs to be under the wings of Rackspace for sometime and there is no need to panic at that point in time.

Ever since Rich Clarke left OpenStack, there was a lingering question on whether Rackspace will hijack the project at some point in time. I spoke with people from both Rackspace and the OpenStack developer community. Rackspace’s only response was “See our track record and trust us based on it”. Even though I am not comfortable with the “Trust Us” slogan from a company whose only aim is to maximize the profits for their shareholders, I don’t see anything in their record that will force me to not take them on face value. The developers (including those who are not on Rackspace payroll) seem to be pretty comfortable with the current arrangement. After attending the event, I also don’t see any reason to disrupt the existing arrangement. With more and more companies joining the partner ecosystem and the Apache license for the code, there are no immediate dangers to the open source spirit/objective. Since the project is still in the early stages, Rackspace’s leadership may even be critical for the long term success of the project. Having said that, I would personally welcome any move that makes the project truly independent of Rackspace

It is a dramatic change in their position in six months and I will add some of my thoughts on this here.

First some stats

Before I offer my point of view on this move, let me list out some of the stats put forward by OpenStack project.

Broad community contribution: Diablo milestone 12 features were contributed by developers from 8 different companies.
Number of companies in the ecosystem > 110
Startups spawning in the ecosystem = 2 and more coming
Widespread adoption by larger organizations including MercardoLibre, CERN, Sony Computer Entertainment America, Fidelity, Disney, NeCTAR, etc..

My POV

In my opinion, the foundation is inevitable. There were widespread calls in the community for such a foundation. Even though this pressure could be a reason for this change in heart, I feel that the biggest contributing factor is Rackspace’s realization that they can’t control the project anymore. They may still have majority in the governance body but they are smart enough to realize that the project has crossed the critical threshold beyond which their attempts to control will get negated by either community pressure or forking. Even before anything happened, we saw Dell preemptively trying to stop any attempts by HP to fork the project. By now Rackspace must have realized that there is not much they can do by having OpenStack project under their wraps. Instead, they could put it out in open as an independent entity and attract more participation from the community. With Dell, HP, Citrix and Cisco spending their resources on OpenStack, the chances of Rackspace controlling the project in the future is almost zero. Also, the money thrown by VCs on startups in OpenStack ecosystem means that there is even more resistance to any potential Rackspace takeover (Note: I am not saying they would have done that. I am just highlighting concerns).

Let us be brutally honest here. Rackspace is not a non-profit organization and they are here to make money. This move to release their storage code (along with NASA’s compute) as an open source project was a hail-mary pass against Amazon’s lead in the cloud marketplace. It was expected that such an open source project will put roadblocks on the Amazon’s path by spawning more and more cloud service providers and by tempting enterprises to take the private cloud route. In fact, this has worked to a large extent and we are seeing more and more organizations using OpenStack for private clouds and slowly a federated cloud ecosystem is emerging in the horizon.

Under such a reality, it makes no sense for Rackspace to hold control and bear most of the overheads. Intentionally or not, they have helped spawn an open cloud ecosystem that has got legs to go on its own. Even if it needs help, there are other big companies who have no choice but to bear the burden. By offloading the project to an independent foundation, Rackspace can easily let others share the burden and, anyhow, they are going to have a bigger say because of the usual meritocracy dynamics inside such open source projects. I think it is a smart move based on the reality. It not only serves Rackspace well, it is a great thing to happen for OpenStack community. Kudos to Rackspace for doing it in a more decent way and on their own terms. Kudos to rest of the OpenStack community for taking the project to the current levels. Exciting days are ahead for OpenStack.

disclosure: I briefly gave some strategic suggestions to HP Cloud Division members when they were considering the use of OpenStack for their public clouds. No financial transactions were involved and it is the same advice I give any organization or team that plans to use open source projects.