6.7
Table Of Contents
- vSphere Availability
- Contents
- About vSphere Availability
- Business Continuity and Minimizing Downtime
- Creating and Using vSphere HA Clusters
- Providing Fault Tolerance for Virtual Machines
- How Fault Tolerance Works
- Fault Tolerance Use Cases
- Fault Tolerance Requirements, Limits, and Licensing
- Fault Tolerance Interoperability
- Preparing Your Cluster and Hosts for Fault Tolerance
- Using Fault Tolerance
- Best Practices for Fault Tolerance
- Legacy Fault Tolerance
- Troubleshooting Fault Tolerant Virtual Machines
- Hardware Virtualization Not Enabled
- Compatible Hosts Not Available for Secondary VM
- Secondary VM on Overcommitted Host Degrades Performance of Primary VM
- Increased Network Latency Observed in FT Virtual Machines
- Some Hosts Are Overloaded with FT Virtual Machines
- Losing Access to FT Metadata Datastore
- Turning On vSphere FT for Powered-On VM Fails
- FT Virtual Machines not Placed or Evacuated by vSphere DRS
- Fault Tolerant Virtual Machine Failovers
- vCenter High Availability
- Plan the vCenter HA Deployment
- Configure the Network
- Configure vCenter HA With the Basic Option
- Configure vCenter HA With the Advanced Option
- Manage the vCenter HA Configuration
- Set Up SNMP Traps
- Set Up Your Environment to Use Custom Certificates
- Manage vCenter HA SSH Keys
- Initiate a vCenter HA Failover
- Edit the vCenter HA Cluster Configuration
- Perform Backup and Restore Operations
- Remove a vCenter HA Configuration
- Reboot All vCenter HA Nodes
- Change the Appliance Environment
- Collecting Support Bundles for a vCenter HA Node
- Troubleshoot Your vCenter HA Environment
- Patching a vCenter High Availability Environment
- Using Microsoft Clustering Service for vCenter Server on Windows High Availability
Cause
A vCenter HA failover might not succeed for these reasons.
n
The Witness node becomes unavailable while the Passive node is trying to assume the role of the
Active node.
n
An appliance state synchronization issue between the nodes exists.
Solution
You recover from this issue as follows.
1 If the Active node recovers from the failure, it becomes the Active node again.
2 If the Witness node recovers from the failure, follow these steps.
a Log in to the Passive node through the Virtual Machine Console.
b To enable the Bash shell, enter shell at the appliancesh prompt.
c Run the following command.
vcha-reset-primary
d Reboot the Passive node.
3 If both Active node and Witness node cannot recover, you can force the Passive node to become a
standalone vCenter Server Appliance.
a Delete the Active node and Witness node virtual machines.
b Log in to the Passive node through the Virtual Machine Console.
c To enable the Bash shell, enter shell at the appliancesh prompt.
d Run the following command.
destroy-vcha
e Reboot the Passive node.
VMware vCenter
®
HA Alarms and Events
If a vCenter HA cluster is in a degraded state, alarms and events show errors.
Problem
Table 4‑4. The following events will raise VCHA health alarm in vpxd:
Event Name Event Description Event Type Category
vCenter HA cluster state is
currently healthy
vCenter HA cluster state is
currently healthy
com.vmware.vcha.cluster.stat
e.healthy
info
vCenter HA cluster state is
currently degraded
vCenter HA cluster state is
currently degraded
com.vmware.vcha.cluster.stat
e.degraded
warning
vSphere Availability
VMware, Inc. 92