The main aim of this project is to design and develop a platform for power- and cost-efficient highavailability
virtualisation of servers. Virtualisation is currently regarded as a "hot topic" for multiple areas within computer
science. Within the scope of this project, it will be used primarily as a tool to improve service security - by
compartementalising services into smaller units - and reliability - by migrating virtual servers when physical
machines fail. One main aspect of this project is to analyse and balance all the advantages of virtualisation with its
respective disadvantages, namely the increased complexity and decreased performance. We aim to develop a
reference architecture and implementation that will allow easy and optimal use of virtualisation especially in small
to medium sized enterprises (SMEs). Those can currently hardly benefit from virtualisation due to cost issues.
Specialised - and therefore expensive - mass storage technology like storage area networks (SAN) and network
attached storage (NAS) will be avoided. Instead, local hard disks and a shared-nothing cluster architecture are used
to synchronise data between mirrored cluster servers over standard networks. The present research project should
also create additional benefit for already established virtualisation solutions in larger enterprises, especially with
regards to security and maintainability.