alex777
Contributor
Contributor

HA claster tuning

Jump to solution

Hi.

There are a few questions about HA of cluster.

1) There are two ESX hosts in a cluster. VM1 works on the first.

Then the first host becomes not accessible. VM1 automatically started on the second host of cluster.

Is it possible so to do what when the first host began

to work VM1 again started on it, and for the second host stopped ?

2) If in HA cluster more than two ESX hosts, and does one of them become not accessible, what from remaining hosts VMs will begin to work on from a saying no host ?

0 Kudos
1 Solution

Accepted Solutions
weinstein5
Immortal
Immortal

1) If you are asking is HA bidirectional yes VM1 can fail either way or if you are talking once the original host returns to service will VM1 automatically fail back to the original host no - if you happening to be running DRS and host2 is overloaded yjam DRS will automatically move vms to balance the liad

2) If you mean when server becomes unaccessable that is has failed then VMs will start up on the remaining hosts - if mean the host is still running but not communicating with the rest of the cluster this is referred to as the node has become isolated - and depending what your isolation response will determine where the vms will run - by default they will power off and get restarted on the remaining nodes of the cluster

If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful

View solution in original post

0 Kudos
2 Replies
weinstein5
Immortal
Immortal

1) If you are asking is HA bidirectional yes VM1 can fail either way or if you are talking once the original host returns to service will VM1 automatically fail back to the original host no - if you happening to be running DRS and host2 is overloaded yjam DRS will automatically move vms to balance the liad

2) If you mean when server becomes unaccessable that is has failed then VMs will start up on the remaining hosts - if mean the host is still running but not communicating with the rest of the cluster this is referred to as the node has become isolated - and depending what your isolation response will determine where the vms will run - by default they will power off and get restarted on the remaining nodes of the cluster

If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful

View solution in original post

0 Kudos
alex777
Contributor
Contributor

Is it possible to do so that during work "Virtual Machine Failure Monitoring" failures VMs was restarting on other ESX host(indicated obviously or the least loaded) of cluster of HA DRS ?

0 Kudos