Skip to main content

How VMware HA Works?

In this article, let’s see How VMware HA Works? VMware High Availability (HA) have been introduced together with the release of vCenter Server In 2003. Exactly the same year, additional features such as VMotion, and Virtual SMP technology also released with VMware vSuite.

What is VMware High Availability (HA): VMware HA provides high availability for virtual machines by pooling them as well as the hosts they stay on right into a cluster. Hosts within the cluster are monitored and in the event of failure, the virtual machines on a failed host are restarted on alternate hosts.

How VMware HA Works?

How VMware HA Works

Above image is a two-host cluster, if the host on the left crashed or had a hardware problem, the VM which works on the left host can be restarted automatically by VMware HA on the host which is on the right. The files which are related to VM’s such as virtual disks (VMDK), a configuration file (VMX) and others files are remains located on the shared storage, the VM’s files located in shared storage do not move anywhere, but the VM begins automatically on the host which is on the right host.

VMware NSX Overview and Prerequisites
VMware vSphere ESX/ESXi vMotion migration fails

How HA protects VM’s: In a vSphere HA cluster, the HA protects VM’s from three types of host failures,
1. Failure – When a host stops functioning.
2. Partition – When a host loses network connectivity with the master host
3. Isolation – When a host becomes network isolated.

What is Master and Slave concept?
Whenever you include a host to a vSphere HA cluster, an agent is uploaded to the host and configured to talk with other agents within the cluster. Each host within the cluster functions as a master host or a slave host. When vSphere HA is enabled for a cluster, all active hosts be involved in an election to find the cluster’s master host. The host that mounts the largest number of datastores comes with an advantage in the election. Only one master host exists per cluster and all other hosts are slave hosts. If the master host fails, is shut down, or is removed from the cluster a new election is held.

The master host confirms in case a host responds to ICMP pings delivered to its management IP addresses. If a master host cannot communicate directly with the agent on the slave host, the slave host doesn’t interact with ICMP pings. When the agent is not issuing heartbeats, it is viewed as failed, then that host’s virtual machines are restarted on alternate hosts.

Difference between NAS vs DAS vs SAN vs Cloud Storage
Best Backup Solution for VMware Infrastructure

Leave a Reply

Your email address will not be published. Required fields are marked *