Cost and quality are two driving factors for open source's role as the bedrock for public cloud computing. Google, Amazon, and other public cloud providers simply can't compete with expensive, proprietary license-burdened infrastructure. They need open source.
As cloud computing matures and moves from public to private clouds, however, we may see enterprises flock to free (as in cost) and open (as in freedom) infrastructure, too.
What would this mean for subscription-based open-source vendors?
It might not be pretty. Tim O'Reilly pointed out nearly two years ago that
almost all of the software stacks running on cloud computing platforms are open source, for the simple reason that proprietary software licenses have no provisions for cloud deployment. Even though open source licenses don't prevent lock-in by cloud providers, they do at least allow developers to deploy their work on the cloud.
The cloud just wants to be free; at least, its infrastructure does. And by this I mean "free" as in cost and "free" as in unencumbered, though some have made compelling arguments that the freedom open source provides is perhaps its most critical selling point.
But I suspect that cloud developers, whether building public or private clouds, simply want to get things done at minimal cost, as Gartner research vice president Brian Prentice argues:
[O]pen source is a necessary component of all organizations' supply chain strategies. It is essentially a way to manage cost and mitigate 3rd party dependencies.
Given this interest in something for nothing (or for very little), will enterprises be willing to pay for the cloud? Absolutely. Companies like Microsoft, which just launched its Azure cloud platform, are banking on the notion that enterprises will pay for the convenience of running their applications in a hosted environment.
Red Hat, for its part,that enterprises will pay for the peace of mind that comes from having their familiar server operating system follow them to the cloud.
I suspect that both Red Hat and Microsoft are correct, but I also wonder if they overestimate CIOs' willingness to pay for a component (the OS) that is increasingly being abstracted away. In a world where cloud management services come to matter more than the underlying OS, at least in CIOs' perceptions, will enterprises gladly pay for Windows or RHEL?
It's an open question, and one that may favor OS vendors that don't charge server subscription/license fees, like Canonical (Ubuntu is the dominant OS for many cloud deployments at Alfresco, Zimbra, and other "new-school" ISVs, as is the primary OS for private cloud vendors like Eucalyptus and VMops, too) or VMware (Cloud OS, which attempts to redefine "operating system" altogether), which can afford to give an OS away to charge for other services/software.
Novell, for its part, could leverage its Microsoft partnership to subsidize cloud deployments even as it has server deployments. So long as Microsoft continues to want to hedge against Red Hat, it will likely want to continue splashing cash on SUSE Linux (and it's a lot of cash).
A December Piper Jaffray report highlights the fact that "next-generation Cloud Computing data centers are NOT running Microsoft Windows; they are increasingly leveraging the compelling economics of open source components." Such economics, however, may favor a new breed of open-source vendor, rather than Red Hat, the dominant open-source vendor.
Personally, I think we'll see plenty of Red Hat (and Microsoft) in cloud computing, but it strikes me that the real money is not in the OS anymore, but rather in services higher up the stack, like cloud management tools. Can these two operating system stalwarts afford to give away the OS? Do they need to?
Your thoughts, please.