Virtual Machine and Host High Availability - Getting Started with Red Hat Enterprise Virtualization - page 104

Virtual machine's high availability
Virtual machine's high availability is one of the key aspects that need to be considered
when someone is hosting their business-critical application on virtualization. High availab-
ility means that a virtual machine will be automatically restarted if it is terminated for a
reason other than intentional shutdown by the administrator from the guest operating sys-
tem console or RHEV Manager admin console. A high availability virtual machine is auto-
matically restarted either on its original host or another host in the cluster.
This is achieved due to the way the RHEV Manager is intelligently designed to constantly
monitor the health of the hosts, storage, hardware failure, and more to restart the failed,
highly available virtual machine without any manual intervention and with minimal inter-
ruption to the service.
Generally, high availability will kick in on the following scenarios:
• When the hardware of the host fails
• When the host loses its connectivity with the storage or network
• When a watchdog card is attached and enabled in virtual machines
An RHEV cluster consists of a group of hosts that you can configure for high availability.
In a cluster, you can configure the cluster policy to be in either the load-balancing or
power-saving mode. Under either mode, one of the cluster policies is enabled; the virtual
machine will be automatically live migrated based on the overall load of the cluster. You
can also restrict the automatic migration of the virtual machine, keeping the cluster policy
enabled by pinning the virtual machine to a specific host, thereby allowing manual migra-
tion that suits some specific application use cases.
Before enabling the high availability of virtual machines, the host should be configured for
high availability. A highly available host requires a power-management device and its fen-
cing parameters configured. High availability will only work if the fencing succeeds. For
more information on power management, please refer to the High-availability hosting and
Cluster policies sections later in this chapter.
To configure a highly available virtual machine, perform the following steps:
1. Log in to the Administrator portal, and navigate to the Virtual Machine tab.
2. Select the appropriate virtual machine, right-click on it, and select Edit .
3. In the Edit Virtual Machine window, select Show Advanced Options , and move
to the High Availability tab.