High Availability (HA) is an option for Experience Portal and POM to restart critical systems during an ESXi host failure or general failure with the virtual machine.
Virtual servers run on a failed ESXi host experience downtime until the servers start on another host in the HA cluster. The downtime is the time to detect the failure plus the time for the virtual machine to turn on and start working.
Although it is not required, VMware recommends using a secondary NIC for the management network when configuring HA.
The following are the recommended configurations to set up a HA cluster to use with Experience Portal virtual machines:
Configure each MPP with Restart Automatically set to Yes in EPM. To check the MPP setting, go to and click the name of the MPP server.
To enable the HA cluster feature, select Turn On vSphere HA.
Go to cluster settings and do the following:
Select Enable Host Monitoring check box.
Select Enable Admission Control check box.
Configure the admission control policy to support the failure of one ESXi host.
Note:
When doing network maintenance, clear the Enable Host Monitoring check box. Otherwise, the vSphere HA can detect a false failure.
To support vSphere HA failover, reserve one or more ESXi host(s) or a percentage of resources on each host. These settings vary based on your preferences and available ESXi resources.
In the , select Primary EPM.
Set VM Restart Priority to High.
Set Host Isolation Response to Leave powered on.
These settings enable the Primary EPM virtual machine to prioritize resources to reduce downtime during a failure. The MPPs, Auxiliary EPMs, and other components of the Experience Portal infrastructure have the next highest priority.
Go to , select the Primary EPM:
Enable VM and Application Monitoring.
Set VM Monitoring Sensitivity to High.
Set Application Monitoring to Include.
These settings enable the Primary EPM to restart if it does not receive a heartbeat in 30 seconds. You can also define and use a custom Monitoring sensitivity rule.
Go to and select the following:
Check Select any of the cluster datastores taking into account my preferences.
In the Datastores available for Heartbeat window, select the datastore where the Primary EPM VM runs.
Note:
Your configuration varies based on which failover scenarios and available resources to cover an HA failover scenario. For more information, see Create a vSphere HA cluster in VMware documents.