Dear tea
I have configured HA on my cluster, is there any way i can confirm my HA is working fine and it will take action if something goes wrong.
regards
mr Vmware
In fact, HA will be well demonstrated when we simulate any host or VM failure scenario, thats what the purpose of HA.
If you can not simulate failure (due to production env), you can check some tabs in VI client.
1. Click on host (from VI client)>>summary>>vSphere HA state>>here you can know whether host is slave or Master.
2. Click on CLuster>>Summary >>Explore vSphere HA section (plz check each and every link available here)
This should give you idea that HA is working fine.
If you have DRS/DPM also enabled, there is HA interop with DRS & DPM. Ex. When you have 2 host with less load in a DPM cluster(without HA), DPM should recommend to put one of hosts into standby mode. If you enable HA, DPM should exit host from standby mode.
You can also check HA configuration from API or Managed object browser.
If this is a test environment, try re-booting host (which is HA cluster) with some powered on VMs(You can do it from VI client>>Right click host>>reboot). These Powered ON VMs should restart on another available host in the cluster.
thanks for help,
it's not a test environment, it's a production.
regards
Mr Vmware
In fact, HA will be well demonstrated when we simulate any host or VM failure scenario, thats what the purpose of HA.
If you can not simulate failure (due to production env), you can check some tabs in VI client.
1. Click on host (from VI client)>>summary>>vSphere HA state>>here you can know whether host is slave or Master.
2. Click on CLuster>>Summary >>Explore vSphere HA section (plz check each and every link available here)
This should give you idea that HA is working fine.
If you have DRS/DPM also enabled, there is HA interop with DRS & DPM. Ex. When you have 2 host with less load in a DPM cluster(without HA), DPM should recommend to put one of hosts into standby mode. If you enable HA, DPM should exit host from standby mode.
You can also check HA configuration from API or Managed object browser.
Since it's production, and you cant perform a "real" test, you have to assume that the HA state reported is correct.
You could verify logs (and datastores) for heartbeats, but I'm pretty confident that FDM is working correctly if the HA state (as per above) looks good.
Cheers,
Jon
