PROJECT TITLE :

Fast and Reliable Restoration Method of Virtual Resources on OpenStack - 2018

ABSTRACT:

We propose a quick and reliable restoration technique of virtual resources on OpenStack when physical servers or virtual machines are down. Several providers have recently started cloud services, and the employment of OpenStack, that is open supply IaaS software, is increasing. When physical servers are down, there's a fail-over technique using the high-availability cluster software such as Pacemaker to revive virtual resources. But, it takes a very long time to revive all virtual resources. There is conjointly a methodology for monitoring every virtual machine by using Ping or other ways and restoring a virtual machine when it's down. But, data might be destroyed thanks to the double mounts of virtual machines relying on the timing of failures as a result of restoration strategies of failed physical servers and virtual machines are freelance. Thus, we propose a quick and reliable restoration method with a regular means for plural sorts virtual resources. In our method, Pacemaker only detects a physical server failure and notifies a failure to a virtual resource arrangement scheduler, then a virtual resource arrangement scheduler determines multiple physical servers to restore virtual resources and calls OpenStack APIs to rebuild. The virtual resource arrangement scheduler additionally detects virtual machine failures by employing a Libvirt monitoring module and restores virtual machines while not data loss by handling Pacemaker and Libvirt notifications uniformly. We have a tendency to implemented the proposed methodology and showed its effectiveness relating to quick restoration through performance measurements.