Design of a Fault-Tolerance Hypervisor Application

Contact: Wajeb Saab

Background:

Fault-tolerant architectures often utilize replication. To tolerate faults in a component, multiple replicas of this component are used, such that a fault in one can be tolerated.

Owing to their low-cost and small footprint, virtual machines are a good choice for use as replicas in fault-tolerant architectures. When virtual replicas are used, it is the job of the virtual machine manager (hypervisor) to manage the addition of new replicas and removal of faulty replicas.

The goal of this project is to design an application that can be deployed in the hypervisor. This application should monitor the virtual replicas, recover faulty ones, and manage the addition and removal of replicas.

Project Goals:

  • Design of the application
  • Implementation on Xen hypervisor [1]

Required skills:

  • C/C++
  • Linux
  • Fault-Tolerance

References:

[1] http://www.xenproject.org/

Supervisors: Maaz Mashood Mohiuddin, Wajeb Saab