Continuous Disaster Tolerance in the IaaS clouds

The advances of virtualization technology in Data Centres lead to the development of features such as High Availability, Disaster Recovery and Fault Tolerance. However, Infrastructure as a Service (IaaS) computing clouds that span across distributed areas require even a more advanced mechanism. In this paper we present a Disaster Tolerance (DT) enablement solution that covers all layers of IaaS clouds. Technical aspects of implementation are highlighted, detailing the Seven-stage DT algorithm and the enhancements of Qemu-kvm hypervisor, Libvirt virtualization toolkit, Eucalyptus cloud and the Amazon Elastic Compute Cloud (EC2) public API and tools.