Rivalry between the two open source options distracts us from the true point of cloud computing: It's the services, stupid
By David Linthicum
It looks like OpenStack won't be the only open source cloud computing player. Earlier this week, Citrix Systems released its CloudStack software to the Apache Software Foundation as an alternative to OpenStack. Formerly a member of the project, Citrix cited difficulties in making OpenStack work with its technology as a major driver behind the defection.
CloudStack will launch with about 30 technology partners, many of them already involved with OpenStack. CloudStack claims an Amazon Web Services-compatible native API set. OpenStack has an AWS API compatibility feature as well.
[ InfoWorld editor in chief Eric Knorr took an early look at OpenStack's quest to become your data center OS. | Get the no-nonsense explanations and advice you need to take real advantage of cloud computing in InfoWorld editors' 21-page Cloud Computing Deep Dive PDF special report. | Stay up on the cloud with InfoWorld's Cloud Computing Report newsletter. ]
For those expecting an apples-versus-apples fight, you're way off. CloudStack is a software product with an install base that's been moved to open source for marketing purposes. This game has been played many times before, with everything from CRM software to ESBs. Indeed, you'll find open source versions of most major software patterns.
Compared to CloudStack, OpenStack is less finished. Rather, it's a framework or a base set of code that companies such as Piston Cloud Computing and Nebula are able to leverage as a foundation for their products. OpenStack is more complex and not truly baked for the end-user without distribution support. However, OpenStack supports advanced networking capabilities, and it can be implemented for specific use cases, such as high-performance and widely distributed computing. OpenStack also uses the Apache open source license, and the forthcoming Essex release includes contributions from more than 200 developers from 55 companies. Each approach has its advantages and disadvantages.
I don't have a problem with this battle potentially becoming the cloud version of Microsoft versus Apple. But I don't like that the focus has seemed to shift from building and deploying cloud computing services, both private and public, and back to software.
If the goal is to construct better shared cloud services to displace the volumes of hardware and software in enterprises, I'm all for it. However, if this is about arguing over what "cloud software" will be loaded on the new rack of servers, we're headed in the wrong direction. Services, not software, should be the focus of cloud computing.
