“vCenter High Availability” is a new feature in vSphere 6.5 for protecting vCenter server against hardware and software failures by running two instances of vCenter, one is considered as the active node, which is accessed for vCenter management, the other node is a passive node, waiting for the active one to fail to take over and become active.
Learn more information and how to configure, best practices, and important notes about “vCenter HA” here:
How to deploy vCenter HA – page 12, and page 62.
vCenter HA best practices
Important things to know when considering vCenter HA, the following questions are very important to you when planning for vC HA, these questions were answered by VMware technical support team:
1- What makes vCenter HA better than just using the traditional VM FT, i find FT less complicated?
vCenter High Availability (vCenter HA) protects vCenter Server Appliance against host and hardware failures. The active-passive architecture of the solution can also help you reduce downtime significantly when you patch vCenter Server Appliance. On the other hand, FT has a similar functionality, where You can use vSphere Fault Tolerance for your virtual machines to ensure continuity with higher levels of availability and data protection.
You can use vSphere Fault Tolerance (FT) for most mission critical virtual machines. FT provides continuous availability for such a virtual machine by creating and maintaining another VM that is identical and continuously available to replace it in the event of a failover situation.
The reason vCenter HA is apt for vCenter server is that, VC HA does not have requirement of shared storage. Also FT is used for high mission critical applications. But such feature is not necessarily required for vcenter server as its just a management interface and can afford a few minutes of downtime in case of hardware or VM failure before its failover to another instance.
Also vCenter HA does not require separate vmkernal portgroup to be created on the hosts and and incurs additional traffic on the vmnics chosen for FT traffic. vCenter HA on the other hand requires an internal network to be created and just needs a Virtual machine portgroup on the hosts where Active, passive and witness nodes would be placed.
Additionally FT has certain limitations. Please refer to the link below
So all in all, Fault tolerance is an effective feature, but needs to be used wisely for high mission critical applications.
2- What If I wanted to expand the storage capacity? is this going to be reflected to the passive node? or I will have to destroy the cluster and rebuild again?
The short answer is No. The configuration change at the VM level made will not be reflected to the passive node.
3- When I delete logs from the active nodes, are they going to be deleted from the passive one as well?
No. Any deletion or creation of files on active node will not be reflected on to passive node.
4- is it really a MUST to use a different VLAN/subnet for the HA traffic?
Yes. This is one of the requirements of vCenter HA. This network has to be any network other than the Management network. All three nodes ( active/passive/witness) can be on different subnets/networks.
5- is it possible to span two DRS clusters? like having the active and witness in a cluster, and the passive in another cluster? (I have only two hosts in the management cluster and not so easy to add a third host.
No. Spanning across clusters isnt supported at this point. All three hosts on which vCenter HA nodes are going to be placed must be within the same cluster with DRS enabled.
6- Is patching only doable via the bash shell? not the vCenter management interface at 5480?
Patching is possible using both VAMI and bash shell. Refer to the link below.
For vCenter backup and restoration with Veeam, refer to the following guide please