Failover in OAS

I need to install OAS (or a competitor) in such a way that any single machine can be obliterated by an axe-wielding maniac, but the system as a whole will keep on processing requests. While it seems possible, in theory, to set up a multi-node configuration, I'm concerned by the existence of a "primary" node, which smells a lot like a single point of failure.

Short of installing a bunch of machines and getting out my handy axe, does anyone have any experience with this?

Also, it's not clear to me exactly how the web listener can be configured to failover to different request brokers. What protocol is being spoken between the listener and the broker, anyhow?

Thanks in advance...

-----------== Posted via Deja News, The Discussion Network ==---------- http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own Received on Mon Mar 01 1999 - 17:22:00 CST