All i'm trying to find out and so far haven't obviously even though i've looked at the linux-HA website is if:

The HA linux heartbeat clustering stuff, how does it fail over. IE When a problem is detected does one server turn off and another server in the cluster take over the operation of running the services the master was running. If fact perhaps i'm a little off and that the systems ing the HA cluster aren't clones of each other, i would have thought they need to be. As you can tell i'm a little confused here.

Thanks for your guidance to the correct paths

well ha is a daemon that manages ip availability. the only job is to ensure that certain ip addresses are available to a network. the actual sdervices being run on the boxes are irrelevant. ther's certainly no need to maintain mirror images of servers, for example you might have a main box for www, ftp, smtp etc... and run a single backup for all functions in one box due to budget or something. here you'd have 3 ha instances preferring each using an active passive box with the dedicated boxes being the primary in each pair, and the single backup being the secondary. if any box fails, the backup box takes over the ip address of that box and so you stay in business. how the apps on the boxes work and syncrhonize data etc, is your problem, not linux-ha's.

Sorry i just dont understand what you're meaning is on this part?
one other thing, heartbeat as you say doesn't do any syncing, unlesss set-up of course i know it can be used with several different packages like DRDB etc, but doesn't that make heartbeat like the Fake package then?

Thanks for your reply

well these kinds of tools have subtle differences and i'm not directly that familiar with linux-ha in it's own rite, but all any daemon like this is for is load balancing or failing over at the netwrok level.