Virtual Infrastructure Failover Cluster Manager APIs
FailoverClusterManager provides operations to manage a vCenter High Availability Cluster (VCHA Cluster).
A VCHA Cluster consists of three VMs. One is the Active vCenter VM that serves client requests. Second is the Passive VM that is identical to the Active vCenter VM in terms of resources and capabilities. Passive VM constantly receives updates from Active VM and takes over the role of Active vCenter VM in the event of failover. Third is the Witness VM that acts as a quorum VM in a VCHA Cluster. Sole purpose of Witness VM is to avoid classic split-brain problem in a VCHA Cluster. A VCHA Cluster has following states - 1. Healthy - All three nodes in a VCHA Cluster are healthy and connected. State replication between Active and Passive node is working and both nodes are in-sync. 2. Degraded - A VCHA Cluster is said to be in degraded state when it has lost one of the three nodes. Node loss can be due to various reasons and as a result the lost node is not visible to other two nodes. If an Active node is lost, Passive node will take the role of Active node. If Passive or Witness node is lost, Active node will continue to serve requests. A VCHA Cluster can also be in degraded state if state replication fails between Active and Passive nodes. 3. Isolated - All three nodes are isolated from each other. If this happens while VCHA Cluster is in enabled mode, Active node stops serving client requests. If nodes are isolated in a disabled VCHA Cluster mode, Active node continues to serve client requests. A VCHA Cluster has following modes - 1. Enabled - State replication between Active and Passive nodes is enabled and automatic failover happens if Active node fails while the VCHA Cluster is in a healthy state. 2. Disabled - All three nodes are part of VCHA Cluster but state replication and automatic failover is disabled. 3. Maintenance - All three nodes are part of VCHA Cluster but automatic failover is disabled while state replication continues. Active node continues to serve client requests even if Passive and Witness nodes are lost.