Hi All
We aim to re-jig our current config with the 'keep it simple' theory. Here it is, interested to hear your thoughts on the accuracy etc:
ALL 3.5
ESXi hosts:
1 x VMkernel on VLAN123 used for management traffic
1 x VMkernel on VLAN122 used for vmotion traffic
Both port groups attached to single vSwitch
Single vSwitch attached to 2 x physical NICs spread over 2 x physical switches
Port group VLAN123 set to use VMNIC0 as primary and VMNIC1 as standby
Port group VLAN122 set to use VMNIC1 as primary and VMNIC0 as standby
Failback set to YES
das.failuredetectiontime set to 60000 (60 seconds) to allow for switch failover/failback etc
ESX hosts:
1 x Service Console on VLAN123 used for management traffic
1 x VMkernel on VLAN122 used for vmotion traffic
Both port groups attached to single vSwitch
Single vSwitch attached to 2 x physical NICs spread over 2 x physical switches
Port group VLAN123 set to use VMNIC0 as primary and VMNIC1 as standby
Port group VLAN122 set to use VMNIC1 as primary and VMNIC0 as standby
Failback set to YES
das.failuredetectiontime set to 60000 (60 seconds) to allow for switch failover/failback etc
In both of the above configurations das.isolation addresses WILL NOT be added to the HA cluster advanced settings.
My understanding is that in ESX the default service console gateway will be used as the host isolation response address.
My understanding is that in ESXi the NON VMOTION VMkernel will be selected as the gateway source for the default host isolation response address
Thanks
What about shared storage?
This might help too: