No single point of failure

[Rich]

I have 2 CUPS PRINT servers that will need to serve print requests from multiple linux and solaris servers on different subnets.  I am trying to architect a solution now.  My main concern is single point of failure.  My goal is that if one CUPS print server is down, the other CUPS print server can take over.  Is this possible with the Browsing features?  The servers that will print are on separate subnets.   I don't think it is viable to have 1 server on each subnet act as a browsing relay.  Does anyone have any suggestions on how this may be accomplished?  Thanks.