alphaspirit - Fotolia
With vSphere HA, a virtualized server can go down and the system will automatically fail over to a second instance of the same VM on a new server without missing a beat.
First, get VMware cluster nodes set up and running to your usual standards. You need to perform the same updates to all nodes -- two servers in this case -- and have a shared storage space -- your third, storage server -- for the nodes to use. All VMs and their configuration files must reside on, and have access to, shared storage. Without this, when a node goes down and a new node comes up, the data on the new node will not be up to date.
Once you have created a cluster, populate it with hosts and configure vSphere HA settings.
To create a vSphere High Availability cluster, first select the "Hosts & Clusters" view within vSphere. Then, right-click the correct data center to work in (within the inventory tree). Click "New Cluster" and walk through the New Cluster Wizard. Do not enable vSphere HA or the supporting feature Distributed Resource Scheduler yet. Once you have created the cluster, click Finish.
Right-click on the newly created cluster and select "Edit Settings." On the Cluster Features page, click "Turn On vSphere HA." Configure the vSphere HA settings per your requirements. And finally, click "OK" to dismiss the cluster's Settings window.
A single host is automatically elected as the master host when you create an HA cluster in vSphere. The master host monitors the state of the VMs and slave hosts, and communicates with the vSphere server.
To specify behavior, such as the address to determine if a host is isolated from the network, move into the vSphere HA advanced options via the cluster's Settings dialog box.
If you're looking for solid virtualized failover -- small or large scale -- there is no better route. VMware vSphere High Availability is deployed the same for a cluster of two servers and one shared storage box as for a deployment of hundreds of servers.
Do you need VMware HA or Fault Tolerance?
Learn about VMware FT
When vCenter Server's Heartbeat stopped
Dig Deeper on Data center ops, monitoring and management
Related Q&A from Jack Wallen
It's important to prevent data loss during a Linux OS upgrade. How can partitions or disk drives help IT teams meet that goal? Continue Reading
I deleted a partition on my Linux system, but I still see a boot menu. How do I re-install GRUB2 and not lose data? Continue Reading