A recurring question I hear is “how to start a private cloud hosting service project?”. In my opinion, the first step is to know exactly what services the IT organization provides to its users. Then measure the service levels achieved today. And identify the service levels that users expect. This analysis is also important to measure how much it costs for the company to offer such service levels. With such information one can evaluate the impact of implementing a private cloud in these services.
There are several services that may migrate into the clouds with very positive results. A simple example: imagine the process in which development teams seeking physical resources for your project. Suppose a physical server takes 45 to 60 days to be hired and delivered operationally. Timepiece this period, the development team is underused and delivery of the project is offset by the same time period, perhaps even sacrificing some competitive advantage that two months would bring.
Suppose now a private cloud created to meet the developers in a dynamic way in which they request resources (virtual servers) via a self-service portal and provisioning of these resources is that in a few hours instead of two months? Now imagine you are about 200 projects per year … The benefits become clearly tangible.
Other variables can be used to study “business case” used to assess whether the service may or may not be adopted in the cloud hosting services. A better use of servers is one example. The servers used for development environments tend to use a very low, almost going through periods of idleness. An automatic management, enabling provisioning and release these resources greatly increases their use and thus possibly fewer servers (less capital) will be needed. Also it is possible to postpone any extensions to this park (again better application of capital). Certainly CFOs thank!
The productivity of the technical team is a factor that deserves serious attention. We have spoken with the development team, but we must also look at the technical staff responsible for maintaining these servers. How long they spend doing upgrades of hardware and software? And if much of this work is automated? The benefits will also be easily tangibilizar.