redundant squid server concepts

        I've got two boxes running squid (1.1.20) as peers currently
primarily for fail-over redundancy. currently each server configured
to handle the other's IP in the event of failure (which requires a
manual intervention). load "balancing" is handled by which subset/ip
address the client is using/coming from.

        Anyone have a more dynamic method of handling this? Ideally it
would be useful to have a mechanism such as a DNS MX record lookup so
that if one of the squid servers goes down it would fail over to the
other without intervention. I haven't looked at recent versions of
BIND to see if this capability has been incorporated or not (I know
they are working on more dynamics to better deal with DHCP). Currently
if a client is pointed at a specfic cache, they are dead in the water
if that server goes down. i've currently had a failure which is making
me re-think this config.

thanks for any feedback, ideas,

fj..
Received on Mon Jun 22 1998 - 08:24:14 MDT