Yesterday the news about Rackspace buying Anso Labs, a cloud computing consulting firm who played a big role in developing the compute part of OpenStack project, came out and sent the cloud punditry scrambling for an opinion about its impact on OpenStack. Gavin Clarke wrote an article on The Register raising some concerns about Rackspace pushing its agenda through the OpenStack project. These concerns, even though valid to a certain degree, bordered around FUD putting a question mark over the future of OpenStack project.
The deal leaves Rackspace with all but one seat on the project oversight committee.
That will alarm those in the OpenStack community already worried Rackspace has far too much control of a project that’s supposed to be run by and for the community. OpenStack is under the Apache open-source license.
These concerns will received added impetus given the project’s now expanded to include Cisco and Ubuntu maintainer Canonical on the heals of giant Microsoft and others. Cisco and Microsoft will be unwilling to see a single company running a project they’ve now got a vested interest in.
It is unclear whether the governance process will change to compensate for the new order. OpenStack has just finished a project-management revision process that saw the current board and oversight committee formed at the end of 2010.
Joshua McKenty, who is one of the brains behind the OpenStack compute code from the time it was under the NASA wraps, took it upon himself to address these concerns.
I left Anso Labs, and my position as Chief Cloud Architect, shortly before the acquisition by Rackspace. I have retained my seat on the Project Oversight Committee, as the founder and CEO of Piston Cloud Computing, Inc.
Additionally, Vishvananda Ishaya, while an Anso employee and therefore now a Rackspace employee, was elected to the POC by a large majority of OpenStack community members.
Although we continue to be vigilant to any possible problems with the community-driven governance model, so far the only complaints about it have been from outsiders. NTT, for instance, have had substantial influence on the architecture of the project, and have successfully contributed a large number of features (IPv6 support, etc) in the last release. As in any open source project of this size, working code trumps any perceived or proposed authority.
The Anso Labs acquisition is great news for OpenStack – it will allow some of the most core project contributors to continue to focus on what’s best for the platform, with less concern over (and vulnerability to) the vagaries and politics of government subcontracting.
I thought I will take this opportunity to add my own thoughts to the debate.
- The very first time I spoke to OpenStack folks, I was pushing them hard on the role of Rackspace and the lack (at that time) of a proper governance model. They assured me that Rackspace will not drive the direction of the project and a good governance model is waiting to be released. It was eventually released and accepted by all in the community. The fact that companies like Cisco and Canonical have joined the project speaks about the strength of their governance model. There are more than 50 non-Rackspace companies in the community with a sole agenda of making profit. They won’t be part of this community if they sense that the governance model is weak and Rackspace is there with a hidden agenda. In fact, even before I spoke with OpenStack folks for the first time, I had a chance to talk with some of the participating companies including Cloud.com and they were all pretty convinced about the direction of the project.
- I will be naive if I claim that Rackspace has no agenda with the OpenStack project. I see their move as a well calculated hail mary pass done with the expectation that it will completely commoditize at the infrastructure level and push the market competition to adding value on top of the infrastructure. However, I want to point out that the OpenStack community will keep Rackspace at an arm’s length and ensure that the project is not driven by one company’s agenda.
- Moreover, it is evident that FUD mongers have absolutely no clue on the power of a flexible open source license like Apache License. If Rackspace or any other company tries to derail the spirit of a project like OpenStack, this license allows the rest of the participants to fork the project and take it away from the bad boys. My suggestion to these people is that they should take a sabbatical and do some research on the history of open source.
Having been in Open Source evangelism for almost a decade, I have seen much bigger FUDs being unleashed on open source projects by companies disrupted by them and those who have no clue about them. All I can say now is that people should relax, tighten their seat belts and wait for OpenStack to disrupt the Infrastructure Services marketplace, playing a major role in ensuring an open federated cloud ecosystem.
Wow – thank you for actually coming out and saying this. Two reason’s why we haven’t joined OpenStack:
1. Your article’s contents
2. Security isn’t a concern. The word “security” doesn’t even exist on the website… have a search.
Chris,
I don’t know what part of my article made you not join OpenStack but it is immaterial.
I do agree that security is important and if you are keen on putting it in the forefront, I would say you can do it by being part of the project (because it is Open Source and you have the whole source code to plug security in) than by staying out of it. Don’t agree with the excuse.
Krishnan,
Your article wasn’t the reason for us deciding not to join OpenStack – sorry for the confusion. I was implying that reason’s you wrote about were very similar to my reasons.
I was personally (FireHost was a major sponsor) at the OpenStack launch event at OSCON 2010. At that event it was quite evident this wasn’t an open source ecosystem but a single corporate-run initiative. After probing actual hosting companies who were also involved with OpenStack, they felt the same way.
I’m sure OpenStack has it’s purpose if it’s deployed for environments that do not have security concerns (who shouldn’t have security concerns though..) but we have been through the ringer with all virtualization architectures and can say that hands down – VMware wins. Not by a little but by a lot.
I wish the OpenStack initiative the best of success but it needs to be supported and built by the community not by a company which it’s own agenda.
I do agree with you that a project supported by a community than a corporation is a better option. There are no second opinions on that. Fully appreciate the sentiment.
However, I support OpenStack because this is the best open source alternative that is available that will democratize cloud computing. It is vendor driven but it has a flexible license which will allow a motivated community to fork and take it in the right direction. I want an open federated cloud ecosystem than one dominated by handful of vendors like Amazon, Microsoft and Google. OpenStack has the potential to unleash such an ecosystem in spite of Rackspace’s presence. But, again, it is not an ideal situation but a best available compromise.
Hit enter before editing and finishing. Wanted to wrap up that I appreciate your time responding and writing about this topic.
Thanks,
Chris Drake