What is server sprawl and what can I do about it?

Server sprawl is the inefficient or improper provisioning of servers to handle workloads. Server sprawl happens when underutilized servers proliferate across an IT architecture, and are not used efficiently. This can happen in both physical and virtual server provisioning.

One of the most basic ways to address server sprawl is to simply make sure that existing servers are used to their full capacity before adding new servers. Network monitoring tools can provide analytics and feedback on server workloads that can help human decision-makers figure out how to use servers efficiently. In many cases, companies can benefit from shifting some or all of the server workloads to cloud or vendor systems.

Another strategy that has worked well for many companies is to put a request process in place for every new server. This puts a gate on new server implementation, to make sure that server sprawl is minimized. For example, there may be a ticket process that looks carefully at existing networks before adding any new servers to the equation.