HA Newbie

[edawg]

I realize this has been talked about quite a bit, but I still have some confusion on the topic of HA and how hosts respond to outages.  At a basic level I understand you have a master and a slave configuration with two types of monitoring taking place. The network monitoring between the slave and master and the default gateway the monitoring of the datastores via the datastore heartbeat option.  What I am trying to get a better grasp on is what happens when the network monitoring fails, resulting in either an isolated or partitioned situation but the host can still read the datastores.  If the vm option is set to leave powered on will the hosts stay up?  Conversely, if the setting is to power off will they power off even if the host can still talk to the datastores?  Just trying to figure out exactly how hosts respond when the two types of monitoring have different failure scenarios. 

I have read a great article at vSphere High Availability (HA) Technical Deepdive - Yellow Bricks but still seem to be missing the details.

Thank you,

Erik

[rcporto · Accepted solution]

This blog post explain exactly what you're asking: http://www.yellow-bricks.com/2011/10/11/vsphere-5-ha-isolation-response-which-one-to-pick/

Isolation Response is the action HA triggers, per VM, when it is network isolated from the rest of your cluster. Now note the “per VM”, so a host will trigger the configured isolation response per VM, which could be either “power off” or “shutdown”. However before it will trigger the isolation response, and this is new in 5.0, the host will first validate if a master owns the datastore on which the VMs configuration files are stored. If that is not the case then the host will not trigger the isolation response.

To summarize, in your case, if master has access to datastore where VMs resides, HA will trigger the isolation according with per VM settings, that is Leave Powered On and/or Power Off.