Dear All, I am planning on using a squid server as a reverse proxy and have some questions which I would be very grateful if you could answer. The problem that I have is that Apache is currently serving a lot of dynamic content (coldfusion) which hardly ever changes. What I want is to install a fair amount of RAM and have something like squid cache the output of the dynamic scripts so that it perhaps downloads them every hour or so from Apache and then for the rest of the time serves them from either RAM or a disk cache, depending on how popular they are. There are some scripts that are truly dynamic (search, and so on) so I want it to be intelligent enough to pass these straight onto Apache - either through a regular expressing I can define somewhere of URLs not to cache or through some other method (ideally not getting the script to set certain headers...). Is this something that I can do with squid, and if so do you have any hints how to do this based on experience? I am struggling to find a guide to do this (particularly a way to override whatever headers Apache is sending out with regard to expiry date or no-cache and so on), so any help would be appreciated. Apart from setting cache_mem high do you have any other recommendations for increasing/optimising the use of the "hot" cache in RAM? Finally, is it possible to give squid two Apache servers and ask it to use one of them normally but if it fails a health check to switch to another one (other than low TTLs and a DNS switch)? With many thanks, Alex Davies
