Re: [squid-users] Annoying caching problem
Most likely the document was not modified in ages (or clocks are out of synch) when Sqiud originally requested the object. As for how to tune how Squid behaves see the refresh_pattern directive in squid.conf. Regards Henrik On Monday 27 January 2003 15.08, [EMAIL PROTECTED] wrote: > Hi all, > > I am using squid 2.4stable3 on a Red Hat 7.1 box and have a rather > annoying problem. I have documents (i.e. pdf or MS office docs) on > my webserver (apache 1.3.27) which are linked to from my web pages. > So if you klick on the link the dialog box pops up asking you to > save the file to disk or view it with i.e. MS Word. So far so good. > The problem now is that squid always delivers these documents out > of its cache, even if the documents have changed on the webserver. > Looking at a tcpdump, squid only asked the webserver once. After > that, the document is in the cache and squid never queries the > webserver again, so if I update the document on the webserver I > still get the cached version through my browser. > > Now i want to know if this is default behavior. What can I do to > make squid get the copy from the webserver (or at least check for > new versions). I would prefer not having to use the regex match on > document suffixes in squid.conf to tell squid not to cache these > dokuments. Can I somehow include a no-cache in the http header of > the reply, when squid asks for the dokuments? A no-cache pragma in > the web page containing the link to the documents does not help. > > Thanks for any hints > > Marc
Re: [squid-users] Annoying caching problem
[EMAIL PROTECTED] wrote: > > Hi all, > > I am using squid 2.4stable3 on a Red Hat 7.1 box and have a rather > annoying problem. I have documents (i.e. pdf or MS office docs) on my > webserver (apache 1.3.27) which are linked to from my web pages. So if > you klick on the link the dialog box pops up asking you to save the file > to disk or view it with i.e. MS Word. So far so good. The problem now is > that squid always delivers these documents out of its cache, even if the > documents have changed on the webserver. Looking at a tcpdump, squid > only asked the webserver once. After that, the document is in the cache > and squid never queries the webserver again, so if I update the document > on the webserver I still get the cached version through my browser. > > Now i want to know if this is default behavior. What can I do to make > squid get the copy from the webserver (or at least check for new > versions). I would prefer not having to use the regex match on document > suffixes in squid.conf to tell squid not to cache these dokuments. Can I > somehow include a no-cache in the http header of the reply, when squid > asks for the dokuments? A no-cache pragma in the web page containing the > link to the documents does not help. > > Thanks for any hints What does http://www.ircache.net/cgi-bin/cacheability.py have to say for the particular objects. I mean when testing,specify the url containing the pdf and or MS office doc directly, NOT the reference page. M. > > Marc -- 'Time is a consequence of Matter thus General Relativity is a direct consequence of QM (M.E. Mar 2002)
[squid-users] Annoying caching problem
Hi all, I am using squid 2.4stable3 on a Red Hat 7.1 box and have a rather annoying problem. I have documents (i.e. pdf or MS office docs) on my webserver (apache 1.3.27) which are linked to from my web pages. So if you klick on the link the dialog box pops up asking you to save the file to disk or view it with i.e. MS Word. So far so good. The problem now is that squid always delivers these documents out of its cache, even if the documents have changed on the webserver. Looking at a tcpdump, squid only asked the webserver once. After that, the document is in the cache and squid never queries the webserver again, so if I update the document on the webserver I still get the cached version through my browser. Now i want to know if this is default behavior. What can I do to make squid get the copy from the webserver (or at least check for new versions). I would prefer not having to use the regex match on document suffixes in squid.conf to tell squid not to cache these dokuments. Can I somehow include a no-cache in the http header of the reply, when squid asks for the dokuments? A no-cache pragma in the web page containing the link to the documents does not help. Thanks for any hints Marc
