high availability?

Does anyone have any experience implementing Heartbeat or other
high-availability services? I've got a few PII-233s with 2Gb HDs, and
I'd like to make a few network services redundant.

DHCP
DNS
routing (is this possible?)

I know how to make squid redundant, and have a few ideas about DNS,
but I don't have faintest clue of what to do about routing. I have
four or five NATted subnets, and three upstream providers.

Currently I have far too many single points of failure, and zero hot
spares (I know, I know. I'm bad.)

I've seen clustering stuff like openmosix and things like that, but
I'm not trying to cluster processes (I think.) Just provide hot
failover for the critical applications, which on my network are file
serving, printing (needs DNS) and Internet access (with or without
squid.)

(I'd be happy to put my finds into the wiki, Mike.)

Cheers,
Jeremy
---------
To unsubscribe, send email to <aklug-request@aklug.org>
with 'unsubscribe' in the message body.
Received on Mon Oct 18 17:15:36 2004