On 24/06/2014 1:25 p.m., Gregory Morse wrote: > Hey All, > > It's my understanding that squid decides on its own what files to > cache and it does this when multiple different clients request those > files. > > I have an unusual use case, where I want to specify a list of urls > where the files are always cached. This means after the first request > the files on those URLs would be cached by Squid, subject to the HTTP > caching rules of course. > This is the normal operation of Squid. All you need to do is ensure you use a Squid which obeys the HTTP protocol properly. * current Squid releases do HTTP better than older ones, so later is definitely better for this usage. * any configuration option which produces warnings about HTTP violation are to be avoided. Some can ensure caching of certain objects, but only by breaking your criteria of HTTP compliance and not without some form of bad side effect. As Eliezer has already mentioned for reverse-proxy (CDN) adjusting the cacheability controls on the origin server may also be needed to ensure caching (HTTP compliant caching is to do so whenever the server does not prohibit it). RFC 7234 has details on the caching controls if needed. HTH Amos