If Infrastructure as a Service was the cloud buzzword of 2010, and Platform as a Service was the cloud buzzword of 2011, cloud operating system will be the big cloud buzzword of 2012.
Microsoft (MSFT) and VMWare (VMW) are both out with their own cloud operating systems. Microsoft says its System Center 2012 offers the key building blocks for its private cloud strategy. VMWare says its Cloud OS can handle internal clouds, private clouds, and their intersection with public clouds.
Many observers believe OpenStack, an open source project originally sponsored by Rackspace (RAX), may be the cloud OS answer.
Joshua McKenty, Piston Cloud co-founder, explained in an interview with me yesterday that key questions must still be answered before enterprises rush to replace Internet with cloud. Questions like proving your identity to the cloud, authorization to use cloud services, and policy management across clouds.
"We're not going to get hybrid cloud nirvanas until the private cloud has the security and integration businesses need," he said. The development of the cloud is more like that of the Internet, where mission-oriented programmers worked collaboratively for decades, inside private networks, before it all exploded into public view during the mid-1990s.
Today's clouds are more like the Internet of 1988 than that of 1995, he said. These key questions have yet to be answered. McKenty said he based pentOS on the open source Open Stack project, rather than a more proprietary framework like those used by VMWare and Microsoft, because open source is the best environment for seeking answers to those questions.
This doesn't mean you should put all your money on Rackspace, the original sponsor of OpenStack, either, McKenty said. OpenStack is being placed into a foundation, outside Rackspace's control, specifically so these hard questions can be addressed collaboratively.
So what should cloud investors do? McKenty counsels patience. Public clouds like those of Rackspace, Google (GOOG) and Amazon (AMZN) are the best bets right now, because at its heart cloud aims to reduce head count, to increase the number of servers a single system administrator can manage.
For now, costs are the key cloud metric. That will change. Just not in the way stock touts or reporters think it will. Because it will be slow, slower than anyone now expects.
Cloud is hard.