Resolving automatic high availability operations and users commands to virtual machines on OpenStack

摘要

In recent years, server virtualization technology has become more and more popular. Many companies tend to build services on virtual machines (VMs) on cloud computing platforms. Therefore, high-availability (HA) technologies that improve system availability by reducing downtime have become critical. In previous research, we proposed a HA technology called High Availability Service System (HASS) for OpenStack. However, the operation of HASS conflicts with the operation of the VM user or administrator. For example, HASS incorrectly judges a user’s operation (such as live migrating a VM) as a fault and conducts the corresponding failover. There is no existing method that can be applied to OpenStack to solve this problem. Therefore, this paper aims to propose a new method to resolve the conflict between automatic operation of HA technology and VM operation. The proposed method collects VM operation information and automatically changing the fault detection strategy of OpenStack HA technology. According to the experimental results, even if the user operates the VM, the OpenStack HA technology using the proposed method can accurately detect the VM fault and failover.