Michael Jurisch wrote:
Hi there! I am totally new to Squid (have worked/tested it the fifth day now) and I not that good in English language - so have fun. ;-) I want to use Squid 2.6 to mirror certain pages/files of a website. I want the users to get the content out of Squids cache. Therefore I use squid as an accelerator: http_port 80 accel defaultsite=www.foo.de Ok, this is no problem, I get data into the cache and as I resolve www.foo.de to the squids IP I browse through its cache. Now here comes one problem: if I watch in the access log of www.foo.de, there is _always_ a (control) request from squid server to www.foo.de when the browser requests the site (with the http code 304). But I don't want to have this control request everytime, I tought squid requests www.foo.de, not until it recognize, that the file in its cache has expired!? In order to get rid of the control request I configured some refresh rules, e.g. for http: refresh_pattern ^http: 1440 20% 10080
cache_peer real.server.example 80 0 no-query no-digest no-netdb-exchange originserver
That should do it for the ICP queries, and some other potential unwanted ones.
I thought this means, that squid keeps every file that comes via http for 1440 mins in its cache and never looks at www.foo.de until the time is up (I tested with and without the meta tag and different expiration times in the html file). But unfortunately it always does this control request. Do you have any idea how to disable that?
refresh_pattern only works when Expires: or Cache-Control: headers are MISSING.
Another question is whether you know a good file manager to browse and manipulate the cache? I was finally able to get "cachepurger" running (was hard to find, as it seems that www.isp_systems.com is down), but it offers to less and doesn't seem to work really correct (or the behaviour of squid is strange, because when I look at "All Cache Objects" in the "Cache Manager Interface" I see slightly different files).
Haven't looked at this yet myself sorry. Amos