Re: [squid-users] Re: TCP_REFRESH_HIT unwanted

Hi!

> Try using a tool that will show you the http headers from the origin
> server e.g. wget -S

I did it and this is the result:

  HTTP/1.1 200 OK
  Date: Mon, 29 Oct 2007 10:15:58 GMT
  Server: Apache/2.2.4 (Linux/SUSE)
  Last-Modified: Fri, 26 Oct 2007 09:44:31 GMT
  ETag: "4001c-8b-328c7dc0"
  Accept-Ranges: bytes
  Content-Length: 139
  Cache-Control: max-age=300
  Expires: Mon, 29 Oct 2007 10:20:58 GMT
  Keep-Alive: timeout=15, max=100
  Connection: Keep-Alive
  Content-Type: text/html

So, how do I go on now to get rid of the REFRESH? IMHO the headers look good - 5 minutes to keep files fresh. So what could be wrong?

Thanks a lot,
Micha

-------- Original-Nachricht --------
> Datum: Fri, 26 Oct 2007 16:17:12 +0100
> Von: RW <fbsd06@mlists.homeunix.com>
> An: squid-users@squid-cache.org
> Betreff: [squid-users] Re: TCP_REFRESH_HIT unwanted

> On Fri, 26 Oct 2007 14:26:12 +0200
> "Michael Jurisch" <archer@gmx.org> wrote:
>
> > Hi there,
> >
> > I found some small hints that hopefully will lead to solve my problem
> > (http://www.squid-cache.org/mail-archive/squid-users/200710/0260.html).
> >
> > To make it short: in the squid access.log I mostly see
> > TCP_REFRESH_HIT/304, but I want them to be TCP_IMS_HIT. If I see it
> > right, TCP_REFRESH_HIT/304 means, that the object in cache is not
> > considerd as fresh, that's why there goes a request to the origin
> > server, which tells squid, that the object wasn't modified. But I
> > don't know, why the cache thinks, the objects I test with (2 simple
> > html files and one jpg) are stale!? I installed mod_expire for apache
> > and I configured it in a way that the objects should be fresh for 5
> > minutes.
>
> Try using a tool that will show you the http headers from the origin
> server e.g. wget -S
>
>
>
> > To make it clear: I don't hit reload all the time, I just
> > click links that connect the two html files, so I only jump from file
> > 1 to file 2 and back. In my opinion this has to be TCP_IMS_HIT. So
> > what factors influence this? What can I do to get TCP_IMS_HIT instead
> > of TCP_REFRESH_HIT?
>
>
Received on Mon Oct 29 2007 - 02:22:35 MDT