There are two components you are confusing with each other:
The two are not necessarily related.
Central Mangement is the sharing and replication of configurations among master, Sub-master and Site Instances. you logon to the Master and make all your configuration changes. When you apply, the changes are replicated to the other devices. It is used strictly for configruation synch of the policy. It does not care about how proxy traffic is sent to the applaince(s).
High Availability shares a virtual IP address using Heartbeat and is simply a method of creating a Cluster IP address that distributes TCP connections to a proxy that will service the request. It uses multicast a MAC to listen on the Cluster IP address and the proxy that services the request responds and handles the connection.
You can run one, or the other, or both, but they don't relate to the other.
You typically do not need to use High Availability in most environments if you use PAC files that can pick a proxy and send the traffic directly.
Thanks for the response.
I'm aware that Central Mgmt and HA are different, but the docs clearly state that you are advised to set up Central Management first, then configure HA, I guess to make sure that the HA details are pushed to the site instance.
The reason we are using HA is for the proxying of application based traffic which cannot use a Pac File.. i.e. an internally hosted website needs to pull an external data feed.
The question still remains though, are we able to use Unicast to for the HA to function or are we restricted to only Multicast?
Thanks in advance..
unfortunately I am not aware of any (supported) way to re-configure the HA running on MWG 6.x to not use that kind of discovery to communicate to other nodes.
It may be possible to use a seperate NIC for the heartbeat and put both nodes heartbeat NICs to a seperate switch or VLAN to prevent the heartbeat traffic showing up on the network.