Once the HA cluster has been created, the vCenter High Availability maintenance and test failover procedures ensure the VCSA 6.5 is protected against multiple types of hardware and software failures.
When the vCenter Server nodes (Active, Passive and Witness) have been deployed, the maintenance and test failover of the HA cluster should be performed on regular basis to guarantee a working and protected cluster.
VMware vCenter High Availability: setup - pt.1
VMware vCenter High Availability: maintenance and test failover - pt.2
Monitoring the HA cluster status
Checking the vCenter Server HA status on regulare basis, ensures the maximum protection for the virtual infrastructure functionality.
To check the current status of the HA cluster, open the vSphere Web Client and select Configure > vCenter HA. Click vCenter HA Monitoring to access the monitoring window.
The window displays the current status of the Active, Passive and Witness nodes. Click vCenter HA Settings to go back to the management page.
HA cluster maintenance
The HA cluster is enabled by default and performs an automated failover when a failure occurs. If you need to control the behavior of the HA cluster, from the vCenter HA area click on Edit button.
Selecting the Maintenance mode option, the automatic failover is disabled but replication still works between Active and Passive nodes. Click OK to confirm. This option is used during the update procedure of the vCenter Server.
When the HA cluster is in Maintenance Mode, the current Status is displayed in the table.
Selecting Disable vCenter HA option, it disables both failover and replication but HA configuration remains intact. Click OK to confirm.
When the HA cluster is disabled, the current Status is displayed in the table.
Selecting the Remove vCenter HA option the HA cluster is completely removed but the Active node continues to run normally.
If the HA cluster is in Maintenance Mode or Disabled, select Enable vCenter HA to restore the default HA cluster functionality.
When the HA cluster is enabled, it requires some minutes before the automatic failover protection is available again since the intial replication may still be in progress.
The vCenter HA is available again and the automatic failover protection is enabled.
Backup and Restore
As backup strategy, backup the Active node only skipping the Passive and Witness nodes. In case you need to restore the Active node, you must remove the HA cluster configuration, restore and re-create the HA cluster.
To test the failover, from Configure > vCenter HA area click the Initiate Failover button.
Flag the Force the failover... option then click Yes to initiate the failover.
You can view the failover progress from the browser by using, of course, same DNS name or IP address. VMware reports an RTO of 5 minutes, maybe a little bit to high but surely will be improved in future releases.
The Passive node took over the Active role and the HA IP address (in the example 192.168.70.10) remained, as expected, the same.
When the Active node comes back, the normal status is restored.
Configuring the vCenter Server HA ensure an higher level of protection for your virtual environment with no extra licenses required. Testing failover is the good way to make sure that everything works as expected.