Chapter 12: Scaling Out KVM with OpenStack
Being able to virtualize a machine is a big thing, but sometimes, just virtualization is not enough. The problem is how to give individual users tools so that they can virtualize whatever they need, when they need it. If we combine that user-centric approach with virtualization, we are going to end up with a system that needs to be able to do two things: it should be able to connect to KVM as a virtualization mechanism (and not only KVM) and enable users to get their virtual machines running and automatically configured in a self-provisioning environment that's available through a web browser. OpenStack adds one more thing to this since it is completely free and based entirely on open source technologies. Provisioning such a system is a big problem due to its complexity, and in this chapter, we are going to show you – or to be more precise, point you – in the right direction regarding whether you need a system like this...