>I was just told today that one of the thoughts behind running multiple Web
>Server instances on the same server with load balancing software is to avoid
>outage if one of the instances has problem (instance hang, instance
>encountered software bugs/errors, instance brought down unintensionally,
>etc.). Does this make more sense?
It makes some sense, but have you heard what happened about 40
seconds after the launch of the first Ariane 5 rocket on June 4,
1996? The rocket veered off course and exploded. The reason
this happened is that one of its on-board computers had a
software failure. So, the rocket reverted to a backup computer
which was there specifically in case that computer failed. Did
that help? No, because the backup computer was running the same
software as the primary computer, so it failed in the same way.
I'm not saying that this kind of thing will happen to you
if you run a web server on two ports, but I am suggesting
that it's easy to overestimate the gain in reliability you
might get from running two instances of the same software.
- Logan
--
"In order to be prepared to hope in what does not deceive,
we must first lose hope in everything that deceives."
Georges Bernanos