Virtualization refers to the abstraction of computer resources. This technique allows the virtually creation of servers, the installation of software packages on these virtual servers and the running of the installed software packages as if they were located on an actual machine. The technique was primarily used to minimize space allocation necessary to store servers, to reduce electricity power the servers consume and so to save money. Other perspectives are the ability to balance from one configuration to another depending on the current load in a completely transparent manner for the end-user.

The virtualization has progressed so that complete technical environments can be virtually created (servers, networks, sets of application…). On the other hand it becomes more and more difficult to understand where applications and servers are actually located and then to evaluate risks for the business if any technical item fails.

This is the reason why there are more and more requests for virtualization modeling and the ability to deduce impact of virtualized item to the business.

A computer cluster is a group of linked computers, working together closely so that in many respects they form a single computer. The components of a cluster are commonly, but not always, connected to each other through fast local area networks and share common storage means. Clusters are usually deployed to improve performance and/or availability over that of a single computer, while typically being much more cost-effective than single computers of comparable speed or availability.

Since the group of clustered computers is viewed as a single one by the end-user the clustering technique can be considered similar to the virtualization technique. But the slight difference here is all the machines may exist … or be virtual. In this context, the performance improvement is due to the grouped computation capabilities while the virtualization objective is to transport end-user machines to the most appropriate location in real-time.

Clustering leads to the same questions as virtualization does: What are the impacts of a server failure on my business? Is this actual server really useful for my tasks? And so on. To answer such questions, the modeling of clustering is also necessary because the business features is attached to the apparent server while the supporting applications are hosted by the clustered servers.