sön 2008-04-06 klockan 05:15 -0400 skrev Nick Duda: > Do you have an install guide? I'll go check the wiki. There is several out on the Internet, and some in the Wiki. > My thought is that a Cisco device is more stable than a > "server" (proxy server has spinning harddrives...etc). If the proxy > was to die, and all the clients gateway was the proxy we would have to > reconfigure them via DHCP, have then release/renew to a new > gateway...etc (a head ache). If we can do redirection from the cisco > device and the proxy dies, then we just change a setting on the cisco > device to not forward to proxy. Depends on what service level you need. If it's running as a bridge then it's only to bypass the server (simple reconfiguration in the switch if planned right, may even be automated). But yes, the easiest alternative is using WCCP if your Cisco ASA supports that, and from a quick google it looks like ASA supports WCCP. http://www.cisco.com/en/US/docs/security/asa/asa72/configuration/guide/dhcp.html#wp1094445 > I still cant find anyone on squid-users or even smb/winbind/linux > groups that can tell me how to build failover linux servers so if my > proxy dies the other kicks on. www.linux-ha.org is a good start. or if you use WCCP then just have both register with the router and it's automatic.. > We love squid and use it everywhere but it is a major point of failure > that would require more time to fix then other forwarding > solutions....i hope to get this going. It's all a tradeoff between recovery time, configuration complexity etc. One question you should ask yourself is how long downtime is acceptable if the proxy should malfunction, and map this to your support organisation capabilities. Regards Henrik
