Open Source Cloud Stack Community Grows Quickly
April 28, 2011
A coalition of technology companies has come together in just nine months to deliver an open source software stack to enable cloud computing. At the OpenStack Conference and Design Summit going on this week in Santa Clara, Calif., participants in this open source community provided details on two interrelated open source projects: OpenStack Compute for developing a cloud-based server environment and OpenStack Object Storage for cloud-based storage. Project organizers said open source software will allow businesses to use the same cloud platform in their own environment or with an external cloud service provider.
Initially formed early last year by cloud service provider Rackspace, OpenStack last year joined forces with the NASA Ames Research Center, which was working on its own open source cloud platform. Rackspace is behind the storage OpenStack while NASA is behind the compute OpenStack.
In that short time, OpenStack has already put out three versions of its cloud stack, the latest on April 15, while a fourth is scheduled for release this summer. The number of developers contributing to the project has grown to 70 from five a year ago. The community has drawn support from 50 organizations, double the number last year, including corporations such as Cisco Systems, Dell, Intel and NetApp, as well as a number of cloud start-ups.
"There was a vacuum in which an open source project needed to exist," says John Engates, chief technology officer of Rackspace. Cloud computing technology available today is a proprietary offering, he adds. from cloud service providers like Rackspace or Amazon Web Services or cloud software platform vendors like Microsoft.
Other big companies, such as IBM, host a customer's private cloud in its data centers. But Amazon cloud software runs only in Amazon's public cloud, and Microsoft's cloud programs, such as Azure, run only in Microsoft's data centers. "With OpenStack, that [cloud] can be run anywhere--in a Rackspace data center, in a competitor's data center or it can run in your own data center," he says.