I’m at OSCON this week drinking from the open source fountain that made RightScale possible. In talking to Tim O’Reilly I noticed that he hadn’t realized how integral open source is to the cloud, so maybe this fact isn’t as obvious as I thought.
Cloud computing is all about the flexibility to launch and terminate servers on demand, or more generally, to acquire and release resources on demand. This can help solve many tricky problems, including reliability, scaling, development, testing, and business flexibility needs. Where open source comes into the picture is when you think about the software licenses for the software stacks you’re running on all the servers you’re launching. If you are normally running two servers but today you need 10, did you consider whether you have licenses for all the software on the additional eight servers? Most commercial software seems to be licensed by the server or by the CPU, and obviously this just doesn’t cut it in the cloud. If it weren’t for open source stacks, no production service would be operating in the cloud today; everyone would still be waiting for software vendors to get it and change their licenses to enable efficient use in the cloud (yeah, right).
But all this is starting to change. The vast majority of software vendors we talk to are in the process of trying to figure out how they can sell their software in the cloud, including what technical changes are necessary to enable their customers to deploy their software into the cloud environment and what business model changes are necessary to offer frictionless sales into the cloud. Of course deploying software on the RightScale platform offers a number of benefits, including some new features we’re currently adding to support publishing and charging by the use. But the bottom line really is that without open source we wouldn’t have cloud computing today.