1 of 1 people found this helpful
I believe they call it multi-process mode.
Currently Web Gateway 6.x only works in a active/active fashion. Both nodes see all of the traffic, the heartbeart tells the node when which node should respond to the request. If you want to setup a failover scenario, this could be done with DNS, or external load balancer, it is not currently that is built-in.
To comment on what nsgmike said, "Multi-Process" mode is available for appliances with sufficent ram (WW1900 or higher). This essentially triples the number of internal ICAP servers (filtering processes) available to the proxy on the Web Gateway. This does NOT give failover capabilities in case one appliance goes down. For more information on the multi-process mode you can read more about it in the System Configuration Administration Guide, it can be found under Home > Manuals in the Web Gateway's (Webwasher) interface, info on MP-mode is located on page 149. Alternativley it can be found in our KB at https://kc.mcafee.com/corporate/index?page=content&id=PD21824.
So a better way to set these up would be in a Cluster, assigning the cluster IP as the proxy IP/default gateway, and allow the load balancing determine the path/device to use?
I have two web gateway appliances 6.8.6 version running in multi-process mode. I have lost one of them before and all my traffic went through the other one no problem. I had no reports of users experiencing any issues and I didnt even know until I received my email alert that one was down.
Correct, you would set them up as a load sharing cluster and assign the cluster IP as the proxy for the clients.
You have two appliances, that is why the failover worked. The MP mode just makes it so each appliance can handle more traffic.