One of the dangers of hosting virtual private servers is what is known as virtual server sprawl. Like constructing buildings in an area, only so many virtual machines can occupy a single server. High demand for VPS accounts may quickly out pace your ability to create new machines or even deploy new servers to host more virtual machines.
For this reason, it is important to have a tight cap on virtual machine resources for every server. Not only should there be a numerical limit for the number of machines, but that limit should also be within the limits of the server’s memory and CPU capacity. Rather than leaving virtual machine capping in the hands of various system managers, you should implement a system-wide policy that is enforced by the management software itself.
It is easy to setup and run a VPS, and customers usually expect them to go live in a short amount of time. Therefore, if you know you will be running low, you should prepare deployment with high demand in mind and be ready to open up a new “block” of development, if we are to continue the construction analogy.
If virtual machines are used internally, they should have life cycles so that those no longer in use do not stay running on a server, taking up valuable resources. With policies for all virtual server management in place, you should be able to enjoy the benefits while also being prepared for any setbacks.