Abstract:
Business applications running on IT
infrastructure necessitate high levels of availability in
order to minimize the amount of downtime
experienced during any planned and unplanned
outages. As a result, disaster recovery has gained
great significance in IT. Exploiting virtualization and
ability to automatically reinstall a host, where the
action on a virtual machine is performed only when a
disaster occurs. Virtualization affords significant cost
and performance advantages over more traditional
disaster recovery options such as tape backup or
imaging. Our approach is to design and implement a
continual migration strategy for virtual machines to
achieve automatic failure recovery. By continually
and transparently propagating virtual machine’s state
to a backup host via live migration techniques, trivial
applications encapsulated in the virtual machine can
be recovered from hardware failures with minimal
downtime while no modifications are required.
Moreover, our framework intends to monitor virtual
machines for problems such as CPU utilization, I/O
activity, and memory utilization. This raises a
difficult problem, since it is quite difficult to
discriminate based on these measures between a
virtual object that is performing properly, and one
that is quite ill. We apply the out-of-band monitoring
using virtualization and machine learning can
accurately identify faults in the guest OS, while
avoiding the many pitfalls associated with in-band
monitoring.
