Until now, OpenStack has appeared in the marketplace as a set of components in major Linux distributions. The Nebula One hardware appliance aims to simplify private cloud.
8 Great Cloud Storage Services
(click image for larger view and for slideshow)
A new hardware version of OpenStack is now available from Nebula, the company that spun out of the NASA Nebula cloud under Chris Kemp, the former CIO of the federal agency.
The Nebula One cloud controller is a single product rather than a set of loosely associated software components. Until now, OpenStack has appeared in the marketplace as a set of components included in major Linux distributions from Red Hat, Suse or Ubuntu. The Nebula One is instead a hardware appliance that fits into the 2u slot of a rack. The customer simply selects a standard two-CPU server of choice, plugs the device into the slot of the controller, and the OpenStack cloud comes to life on the rack. Like the controller itself, the server slots in the rack may be equipped with servers of the customer's choice.
Although some assembly is involved, little has been left for the customer to decide. The Nebula One's combination of server, storage and memory is largely preordained. Server choices are narrowed down to either an HP DL 380p Gen 8, an IBM X series model still being certified, or a Dell r720xd. The three selections are similarly configured x86 servers and each of them is a 2u addition to the rack, like the Nebula One controller.
That's by design. Although Nebula doesn't want to tell customers which servers to buy, it does want the servers to fit into a standard enclosure and to function in a way that matches the capacities of its OpenStack cloud controller. The controller runs preconfigured OpenStack software and in effect becomes the cloud services manager for its rack unit. "It is a turnkey device," Kemp said. "The goal is not to get customers talking about building the private cloud. The goal is to get them using the private cloud."
Until now, cloud computing has been built by large Internet companies with the budget and need to create automated, uniform x86 environments. By installing Nebula One, Kemp said, a Fortune 500 company with a workload mix that doesn't look anything like the highly digitized services of a Google, eBay or Yahoo can have a little piece of Amazon right there in its own data center.
Nebula will tout the advantages of private cloud, such as eliminating the need for sensitive data to leave the premises, keeping operations behind a familiar firewall and avoiding the latencies that often result from operating from a remote public cloud data center. Some companies are already attempting to build out their own version of a cloud.
Some of these efforts are in their second or third year. "Each year, it's a check-box item: 'Build private cloud,' and each year the company finds it's starting over," Kemp said. When OpenStack is brought into a company as software, it might start out as the same system each place it's implemented, but soon one copy is modified one way, another a different way. Discussions over hardware lead to different hardware purchase decisions, and eventually one branch of the private cloud no longer resembles another to which it is supposed to march in lockstep.
Nebula has taken on the task of managing OpenStack and offering a hardened and tested version on its cloud controller units. Having the cloud software on a locked-down piece of hardware also makes for more secure operations, according to Kemp. "We're trying to give the employees at enterprises the same kind of compute power that the [Google, Yahoo, eBay] guys in Silicon Valley have. If they need a scale-out database to work with big data, they should have one," Kemp said.
Nebula One runs the firm's own version of OpenStack, dubbed Cosmos. Unlike some versions, Cosmos is able to recognize Amazon Web Services APIs as well as OpenStack's, which opens up AWS's tools to Nebula users. Kemp said he didn't see customers making use of the APIs set for interoperability so much as sharing tools and other resources found across the two cloud communities.
Small and midsized businesses would probably find sufficient compute power in the servers that could be loaded into a single rack. Larger organizations can add racks for a larger system that keeps three cloud controllers working together for a mutually backed-up cloud system.
A Nebula One cloud may consist of a single server configuration or up to 20 dual CPU servers in a rack. With the simplest configurations of one or three servers, customers can get started with cloud computing for about $100,000, said Chris Kemp, Nebula's CEO.
Kemp emphasized that Nebula is a systems company -- meaning a combination of hardware and software -- not a software company. "We're the first computer systems company to emerge in the cloud era," he said.
As a test bed for its new product, Nebula has provided Nebula One cloud controllers to Xerox Parc, Genentech and the NASA Jet Propulsion Laboratory. All three companies are within a narrow radius of Nebula's Palo Alto, Calif., headquarters.
Since its launch in 2011, Nebula has grown its staff to include Jon Mittelhauser, co-author of Mosaic and founding engineer of Netscape Communications; and Jesse Andrews, a founding contributor to the OpenStack project and a member of the OpenStack Project Policy Board since its inception. Andrews is VP of product development.
Nebula expects to remain a strong contributor to the OpenStack project. It will update the software in its cloud controller frequently and provide updates to customers as often as they wish.
Attend Interop Las Vegas May 6-10 and learn the emerging trends in information risk management and security. Use Priority Code MPIWK by March 22 to save an additional $200 off the early bird discount on All Access and Conference Passes. Join us in Las Vegas for access to 125+ workshops and conference classes, 300+ exhibiting companies, and the latest technology. Register today!
Multicloud Infrastructure & Application ManagementEnterprise cloud adoption has evolved to the point where hybrid public/private cloud designs and use of multiple providers is common. Who among us has mastered provisioning resources in different clouds; allocating the right resources to each application; assigning applications to the "best" cloud provider based on performance or reliability requirements.