Today’s announcement of OpenStack, which is detailed here, involved partnerships with NASA and 25 other companies ranging from household names to smaller software companies but the details about the role the partners will play over the coming months are tough to come by. Those few companies that have issued statements cloak their points in generalities about the larger implications for interoperability, standardization, and a more open model for cloud environments.
Among those who were part of the announcement was Cloud.com—one of the few willing to provide some details about the nature of their involvement. While the details are still shrouded in mystery, a conversation with them did shed light on the code itself and its current applicability, among other things.
When Cloud.com launched with its new domain name (do I really need to link it?) this year during its rebranding effort, I spoke with the CEO for a short time to get a scope of the company and what its focus would be. It was clear then that they were aiming to be one of the frontrunners in the turnkey IaaS market, but at the time, it seemed that this was a company that was dedicated to sticking with the small and mid-sized enterprises. Most of their news has landed on the commercial end of the spectrum and hasn’t been covered much here at HPC in the Cloud, but we have been paying attention to the company.
I caught up with them today after seeing their name appear in the list of partners for the hot-off-the-press announcement from Rackspace that it would be opening up its code and collaborating with NASA to get it off the ground this coming year, much to the delight of cloud standards and interoperability folks. RightScale, AMD, Intel, and a host of others were also mentioned but it was nearly impossible to understand what role any of them would be playing.
I made it clear during the email interview that they’d been on my radar but they seemed focused on the small and medium-sized business market rather than HPC or large scale enterprises for the core of their business. Peder Ulander, CMO of Cloud.com responded that they are not simply concerned with this market but that they also see potential (and clients) in the HPC space. As Ulander wrote, “One of our big advantages is the fact that we architected for scale …which is beneficial to large service providers (i.e. Tata) as well as companies seeking to do large scale cloud computing deployments. Applications like Cloudera for big data analytics run on top of the Cloud.com platform. We view HPC as a market that we fit into as we have architected for scale, however we will work with partners to solidify the solution offering specific to HPC.”
While it seems that this means that they can work with clients that fit into the big data analytics realm in beyond, it takes some partnerships to pull off. Nothing wrong with that, certainly, but I suppose it’s important to place Cloud.com in the context of their competition for reference’s sake.
Despite a lack of concrete focus on HPC, Ulander did express some excitement about Cloud.com being a partner with NASA and Rackspace. He stated that for those coming from the world of HPC, this news is meaningful. In his words, “One of the largest HPC deployments from a cloud perspective has been developed and delivered by the Nebula team at NASA. The engagement of this team, coupled with their contribution of code, brings some of the most advanced HPC/Cloud code being used in deployment today and makes it available to users worldwide. For HPC customers and large scale enterprises, this should help alleviate the concern around the scalability, manageability and security in deploying a cloud for their HPC functions.”
Whether for HPC or small enterprise users, there is no doubt that today’s Rackspace announcement bodes changes for the industry. The only remaining issue is what the partners, aside from Rackspace, will be contributing. Before trying to get an answer to what Cloud.com was throwing into the open source brew, I asked Ulander how their involvement came about in the first place. His response was free from specifics, noting only that Rackspace understood their position as a provider of open source cloud solutions for large scale productions, hence they contacted them at some point in the unspecified past. Like the other partners, some of their folks went to Austin for the workshop that cemented the release and finalized details of partnership agreements.
On this topic, Ulander noted,
“First and foremost, the initiative was kicked off by our customers and partners in the cloud. Any effort to bring together leaders with like minds to help drive open standards and interoperability in the cloud is critical to our success and important that we participate in. This is all goodness for the members, the market and most importantly the customers. As a founding member, we have committed to engage with the community, collaborate on standards projects and contribute code that enhances the community efforts. As this effort spins up, we will work with the other members to prioritize how and where we can contribute.”
He continued about the status of the release and their plans to continue as partners—and more important, how they expect to see the initial use of the open source project. In Ulander’s words:
“From a code/project perspective, the project itself is geared towards highly technical institutions with resources to put on integrating and testing their own solution. Two prominent quotes on the organization’s wiki declare that “This is not yet code that comes with certification” and “OpenStack is probably not something that the average business would consider deploying themselves yet.” These comments are designed to help set expectations that this is not commercial grade, turnkey cloud technology but more something that needs significant investment to get up and running. This is where it is important to take note that we are the only cloud orchestration layer included to take this technology commercial. By being part of this initiative, we are in a great position to engage these customers at technical and strategy level which ultimately brings them closer to launching cloud service on Cloud.com product.
The bottom line is that even the most technically savvy customers are not going to invest in areas of non-differentiation (bottom of the stack) and look to partners to integrate, test and support. This is our role in this initiative and our effort to collaborate, contribute and participate will only further cement our position as the ideal partner for a cloud orchestration solution. We’ve already received some positive engagements and interest from large customers who are also part of the OpenStack initiative as a result of a summit around this project held last week in Austin.”
There is a great deal of vagueness to the partner statements about their roles in the OpenStack project but time should reveal more about how each partner is leading expertise to the project. If anyone has any ramblings on why the details of an open source project like this are shrouded in mystery and vagueness, I would be thrilled to chew on your conspiracy theories.
Oh, and as a side note: My biggest question for Cloud.com has always been along the lines of “what small countries have a GDP roughly equivalent to what you forked over to score that domain name” but that one never gets answered. If anyone ever finds out, I’d be tickled to know—just for my own special sense of wonderment.