I have an ESX 3.5 cluster with HA, Vmotion & DRS enabled.Its a 2 node host.
I am able to migrate powered on Virtual Machines between hosts without them restart during migration.
When I set one of the hosts to maintenance mode, the DRS feature automatically migrates the VM to other hosts.
The question here is
When a powered down a node of this cluster while its in operational state running VM in powered on state, all the VMs gets restarted on other node , instead of being migrated in powered on state.
Is there any reason why VMs restarted when a hosts is powered off in a cluster.
Shouldnt all the VM be up and running in a cluster scenario
If I'm reading your message correctly, this is normal behaviour.
HA is used to recover from a failed or isolated ESX host which requires all virtual machines running on this server to be powered down & restarted on alternative nodes.
This is normal behavior - a major misconception about HA is that it utilizes vmotion - vmotion is only used for DRS not HA. vmotion requires running ESX Hosts on either end but with HA one host has died so vmotion is not possible.