NASA has stopped developing code for OpenStack. This was expected at some point. NASA nor Rackspace are software development houses. Neither has expertise in developing what they call in the government space COTS (Common Off the Shelf) software products. I believe this in part explains the long lead time in OpenStack development.
Now that IBM is involved, hopefully the project will move closer to being a shrink wrap type product that’s packaged and more of a point and click installation or at least scripted. Currently the installation process is a pretty typical experience for open source software. Environments that utilize Linux machines as mainly appliances with no to little Linux expertise may find the installation a little intimidating.
I’m (and I believe OptenStack) more interested in the entry of Google and Microsoft into IaaS as a challenge to the adoption of OpenStack. Microsoft has had time to mature its virtualization management platform and I believe it has even more time before OpenStack releases a generally available full featured enterprise platform. MS wants to position their System Center 2012 Cloud & Datacenter Management product as a cloud management solution and I believe a gateway to their new cloud solution. I haven’t looked at the features yet but Microsoft does a good job of selling yet to be released products and features. HP, Rackspace and smaller players offer an OpenStack cloud but if you want to roll your own private cloud by installing the management infrastructure (OpenStack) on your own existing hardware solution you are currently on hold or need to invest in development which fragments your installation from the main project.
That’s where I believe OpenStack is at risk of VMware, Google and Microsoft taking some steam out of the OpenStack train. In the case of Microsoft, the ability for an organization to take their Hype-V infrastructure and convert them to private clouds is compelling. Specifically, the ability to burst to public clouds using a tool and interface they are already familiar with sounds a lot like the VMware strategy. OpenStack makes a compelling argument for vendor agnostic clouds. But, how much value is in a vendor agnostic cloud vs. the convenience of an integrated stack? How does this effect Rackspace’s approach to marketing Openstack to the enterprise as a hybrid solution?
I’m anxious to kick the tires on OpenStack but the current build’s product features, interface and installation aren’t where I’d like them to be for a production environment. What’s your take on the cloud management race?