Am trying hard to think how to implement the best and easiest failover for my current hosting servers.

w/o running a cluster, i.e. using heartbeat / mon / lvs and stuff (all which are difficult to implement for me level)

can anyone help with coming up with a failover solution.

e.g. my pridns fail, sec dns can take over.
mail fail, backup mx can take over.

however, if apache server fail, what happens?

using round robin dns load balancing doesn't help here since dns will still continue to point to the dead webserver.


:bawling: anyone?

DNS tweaks and FailOver = not good.

You need either Load Balancer (Web switch called in da books) in front of your farm (either dedicated Linux or some network gear specifically built for LBs like BIG-IP).

Good luck.

If you just need failover (not load balancing), you can have the boxes hand a 'service IP' back and forth as they take over for each other. That said, if you don't have some type of heartbeat, power switching, or watchdog system in place (to force a failover), and you have no way to keep content in sync, you really have nothing more than a false sense of security.

A single, quality box will serve you better than making two boxes play nice in some half-baked, cobbled-together way.