April 16, 2012, 10:20 AM — In what will be one of the first large scale OpenStack commercial deployments, hosting company Rackspace is moving its server and other hosting services to the OpenStack API (application programming interface).
Starting May 1, Rackspace customers will be offered default access to Rackspace servers through the OpenStack API, or through a new console that uses the Rackspace API, the company said Monday. Prior to this announcement, Rackspace only offered file access through OpenStack.
The company is also developing a number of additional OpenStack-based hosted services as well, offering storage, networking and databases. Collectively, these services will be marketed as Rackspace Cloud.
The new OpenStack API is "more extensible and more robust" than the previous interface that Rackspace offered its customers, said Mark Interrante, vice president of products at Rackspace.
For the OpenStack project, Rackspace's deployment comes at a crucial juncture. The project's leaders are starting a not-for-profit foundation to manage the open source software stack, more details about which they plan to reveal at the OpenStack spring conference being held in San Francisco this week.
The project just released the fifth version of the software stack, codenamed Essex, which will be used as the basis for Rackspace's offering.
Pricing for the Cloud Servers service will remain the same, with the new interface, starting at $0.015 per hour for a Linux virtual server with 10GB of disk space and 256MB of RAM. Pricing for Microsoft Windows servers begins at $0.08 an hour.
Interrante claimed that using OpenStack can cut by half the time administrators spend commissioning and maintaining their Rackspace servers Rackspace, though he acknowledged that Rackspace hasn't done "an end-to-end study" measuring performance times because the service isn't in production yet. But the potential exists. "It is much easier to change [a setting] in a few hundred servers using the OpenStack API," he said.
Using the Rackspace console, administrators can tag certain sets of servers -- identifying them as production or development servers for instance -- and then apply actions to that specific tag. The interface is dynamically updated, so administrators don't have to refresh the console pages when changes occur.