Looking at OpenStack, a Rackspace and NASA Initiative
In case you missed it, there was pretty big news this week coming out of Rackspace - one of our supported cloud storage partners - and NASA. They announced the release and open-sourcing of their cloud backend, now named “OpenStack”.
From the press release:
“We are founding the OpenStack initiative to help drive industry standards, prevent vendor lock-in and generally increase the velocity of innovation in cloud technologies," said Lew Moorman, President, Cloud and CSO at Rackspace. "We are proud to have NASA's support in this effort. Its Nebula Cloud Platform is a tremendous boost to the OpenStack community. We expect ongoing collaboration with NASA and the rest of the community to drive more-rapid cloud adoption and innovation, in the private and public spheres.”
For those of us here at Nasuni this is a pretty interesting development. Since day one, we have been “cloud neutral.” Our goal is to support many cloud storage systems and providers without the customer needing to know anything about the cloud provider’s API, architecture, etc.
We’ve been successful: Our supported cloud storage backends include Amazon S3, Rackspace Cloudfiles, Iron Mountain, and Nirvanix SDN. (Plus more to come.) The Filer seamlessly communicates with these clouds, adapting to each individual cloud’s APIs with zero effort or knowledge required on the part of the user.
We’ve also seen - and written about – the fact that many people are concerned about vendor lock-in. This is what happens when you or your company invest a large amount of resources – such as time, storage or code - leveraging and adapting to a new system that isn’t a standard. Historically the storage hardware vendors locked you in by using custom components which only they supplied. In the cloud, people assume that the basic concept of lock-in applies to the APIs used to communicate with the cloud storage provider.
This is a valid and very real concern. While we at Nasuni have spent many engineering hours making a product that seamlessly supports all of these clouds, not a lot of other companies or users are going to have the time or engineering talent required to adapt to new or differing APIs.
That’s why we welcome the OpenStack initiative. The release of this code, and the potential to enable an all new generation of cloud storage providers, unified APIs, and common “stacks,” is powerful. It is compelling for those companies who have the resources to run a private cloud or for VARs looking to build out their own clouds. Despite the common APIs that clouds built on this technology will have, we’re confident that the Nasuni Filer will continue to act as the perfect gateway to these smaller clouds.
Our engineers have already downloaded, walked through, and run the code which represents the Rackspace Cloudfiles system (named “swift”). It’s written in the Python programming language - something we’re more than comfortable with - and the code is well laid out and documented. We had it up and running in under an hour.
It’s exciting on many levels to be able to run the code for a large-scale cloud storage system so quickly and easily. What’s more exciting is that the Filer already knows how to talk to the API that has been released with OpenStack through our existing support for Rackspace’s Cloudfiles.
With the impressive list of partners and companies already involved in the OpenStack initiative we’re looking forward to participating where we can. And with a release slated for later this summer, we’re excited about the prospect of supporting other clouds (both internal, and external) that leverage the OpenStack/Rackspace API.
Follow us if you’re looking for continued insights on cloud storage – we will continue tracking the news and weighing in.