Friday marks the three-year anniversary of open source private and public cloud community OpenStack. Formally started by NASA and vendor Rackspace in 2010, the community has expanded to 231 contributing vendors, more than 10,000 community members and 1,000 active contributors. OpenStack has undergone particular growth in the last year, with the formation of a Foundation to direct operations, double the number of its component releases and active community discussions on problems with lock-in and cloud growing pains.
Ahead of this “birthday,” Information-Management.com talked over some of the opportunities and challenges of the initial years and those ahead with John Igoe, who was involved in OpenStack early on at Dell – including the Crowbar initiative – and other endeavors before taking up the role of VP of private cloud at Rackspace.
Information-management.com: Give me the quick takeaway from OpenStack’s perspective and Rackspace’s perspective on the expansive interest in this open source cloud effort.
John Igoe: If you think about where we were initially when we founded the initiative, and all of the inflection points and cloud tipping points in that time, it’s really quite remarkable. The average, in terms of the types of contributions we’re getting on a monthly basis, is impressive: over 200 per month in terms of individual contributions to the code. In terms of a trend for the initiative itself, the demographics have really changed. We started out as a very development-oriented effort and we’ve moved into subsequent phases of adoption by users and adoption by operators, making the community very diverse ... For Rackspace, we believe that “open” is the key word for any hybrid [cloud] environment. The ability to federate clouds, not only ones built by Rackspace but other providers, is going to lead to a very powerful adoption curve by customers.
How has the organization behind the scenes at OpenStack changed over the last three years? Or hasn’t it?
It’s changed quite dramatically. In the early days, the vision and the strategy were carried by the Rackspace team ... and focused on telling the community and the world that this wasn’t a one-off type of initiative. The addition of NASA early on added broad credibility ... because they had [recognized] technical background and engineering to build the cloud. Making their background accessible to individuals was very attractive ... Now, there’s really a misperception that NASA has walked away from OpenStack, but that’s not exactly true. They’re still using OpenStack, still implementing it, but they have other initiatives and their leaders have left to do other startups, many based on OpenStack. But the expansion of leadership has lessened the importance of NASA or Rackspace being visible and vocal over time. Rackspace took a lot of heat early on that there was some underlying motive for OpenStack, that [Rackspace] would never release control. That caused an interesting behavior from Rackspace’s leaders – I remember board meetings while I was at Dell where [Rackspace SVP] Jim Curry would not be very vocal. It was important over time to build a reliable board where control would pass to a group of companies and individuals to take OpenStack to the next level. And that’s what the community members wanted to see. If you look at the way the board looks today, Rackspace is certainly influential as a business and technical contributor, but it’s not the only one ... there are others of equal value ... With the Foundation board up and running, there are strong members ... and leadership members can be more assertive.
In terms of that vision for OpenStack, how has that developed over time? What’s the mission statement now compared with three years ago?
I think the vision now is consistent with what it was early on. It’s evolved from a developer focus to more users and operators. In the early days, it was all about establishing the technology and the features that solved a vision for how cloud computing should work in an open environment. So, you have the initial projects around Nova Compute, Swift for storage, [block storage component] Cinder, Glance for image management ... these projects had to get off the ground and become mature enough to be considered production-class technology. That was a heads-down effort to get people to contribute and establish that vibrant community for small and large projects toward a common goal. In the last year and a half, the actual changes in the demographics have changed the way the community thinks about technology and its business aspects. For instance, Cinder and ... projects around billing and metering are very important with operational capabilities. Operators [who are] actually taking OpenStack and putting it into production want to be able to measure it and know how it’s performing.
Unlike some other open source communities, there’s been no shortage of public talk on the basis of efforts – cloud computing. As this brings more casual operators or a second wave of users into the community, how does OpenStack juggle those introductory needs with the strong, baseline community entering more mature deployments and integration concerns?
The adoption in the market place is being driven by two perimeters. There’s the adoption of OpenStack as an alternative to legacy environments. Customers aren’t afraid of open source anymore. They see it as a viable alternative to the shrink-wrapped product. The questions they have about it are the traditional questions around any deployment: Who’s going to support it? What type of maintenance stream am I going to get? What are the feature suite and ROI around it? When you compare something like OpenStack to a legacy alternative, say in the virtualization space, built by a single vendor, the open environment is becoming more attractive. The other thing is the flexibility. Say you’re a health care organization ... and you have a unique differentiator approach you want to take with your cloud environment, you’re working in open source. There, you can turn to the community or build it out yourself.
Since March, OpenStack has had big announcements of contributions and updates from IBM and Red Hat, among others. You want partnerships and that level of support, but how do you keep too much from being forked off and assuage lock-in worries?
This is a huge concern from the community. At our Boston summit, at the end of my keynote, I challenged the community not to “fork it up.” [Laughs] My message is that we have something really great here, but if someone tries to fork it off and create a proprietary front, we’re really going to have issues. Users are going to drive us in a safe direction. It comes back to the competitiveness of the environment. The board is challenged by the fact that companies are making ... million dollar investments in the OpenStack community, and they have to recoup that somehow. So, the board has to walk a fine line that we’re all cooperating and doing things out in the open, but still creating pockets where we can compete, fairly drive revenue and differentiate ourselves.
What has the backlash been? Given the expansion in both vendors and the community, is there any hard line or number to stick to for maintaining that open cloud mission and integrity?
I’d say that there have definitely been “hallway discussions” and informal chats over the concern across the industry. The growth in the community has been phenomenal. Every summit has been sold out and ... has doubled in growth each year, and that’s one measure of growth that the board is extremely happy with. Now, I don’t speak directly for the board. But another trend we like to point out ... if you look at Rackspace’s code contribution over time, you’ll find that we’re no longer number one. We like that, because it means other members of the community are starting to stand on their own and ... contributing at a rate that’s helping the community. The more companies that commit at the same rate as Rackspace, the more code and better code we’re going to have.
Okay, three years have passed with some pretty substantial milestones and recognitions. Pull out your crystal ball and give me a big-picture view of the next year, and the next three years for OpenStack.
Hybrid cloud computing is going to become a mainstay of the adoption strategy for the enterprise, mid-sized and startup companies. On-premise, public cloud and managed cloud, those combined are going to drive OpenStack adoption. Secondly, for the board, they need to recognize that competition is really good for the industry. We have to figure out how we’re going to encourage that competition. In my mind, I like to say competition is good and communism is bad. We have to make sure we don’t push too far toward “communism” and foster competition in a healthy way. The last trend is that forks will fail. Customers are going to reject forked-off projects and will be looking for companies that are true to an open source, open community environment.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access