RE: [squid-users] always_direct and accelerator mode

Peter,

Hmmm well no_cache *removes* the result from the cache right after it was
served. As you point out, it is used with CGI output for example (And I do
use this feature).

The thing is I want the cache to remain, I just want it "bypassed" based on
the originating IP/subnet. If I use no_cache with an ACL that specifies a
subnet, I'm pretty sure nothing will end up getting cached, since any time
someone internal would request a specific object, it would get dropped from
the cache (For everyone though, not just the people who match the ACL).

So the question remains ... is there anyway to specify a group of computers
that would allways go direct when using squid in httpd_accelerator mode ???

Thanks,
J.F.

-----Original Message-----
From: Peter Smith [mailto:peter.smith@UTSouthwestern.edu]
Sent: Friday, May 17, 2002 4:52 PM
To: Jean-Francois.Doyon@CCRS.NRCan.gc.ca
Cc: squid-users@squid-cache.org
Subject: Re: [squid-users] always_direct and accelerator mode

According to http://squid.visolve.com/squid24s1/contents.htm the
"always_direct" and "never_direct" entries are for parent/sibling
request routing (or overriding.) I think what you are probably looking
for is "no_cache" . For instance, the default "squid.conf" usually
contains the following entry which illustrates this.

hierarchy_stoplist cgi-bin ?
acl QUERY urlpath_regex cgi-bin \?
no_cache deny QUERY

This should work as I have a Squid accel that does not do any caching at
all. I handle this with the following "squid.conf" entry.

# Tell squid to NEVER cache anything:
no_cache deny all

Here are some log entries:
yyy.yyy.yyy.yyy (null) user1 [17/May/2002:04:03:07 -0500] "GET
http://xxx.xxx.xxx.xxx/images/menu1.gif HTTP/1.1" 200 2220 TCP_MISS:DIRECT
yyy.yyy.yyy.yyy (null) user1 [17/May/2002:04:03:07 -0500] "GET
http://xxx.xxx.xxx.xxx/images/1icon.gif HTTP/1.1" 200 563 TCP_MISS:DIRECT
yyy.yyy.yyy.yyy (null) user1 [17/May/2002:04:03:07 -0500] "GET
http://xxx.xxx.xxx.xxx/images/utsw.gif HTTP/1.1" 200 6068 TCP_MISS:DIRECT

Peter

Jean-Francois.Doyon@CCRS.NRCan.gc.ca wrote:

>Hello,
>
>I've just begun using Squid in front of my Zope (A Content Management
>System) setup in order to make things faster, and have been quite happy
with
>the results ... Squid looks like a great piece of software!
>
>My setup is quite simple, one server has Apache/Zope, and another has Squid
>as the accelerator, talking to Apache/Zope on one end, and to the public's
>browsers on the other.
>
>Now the thing is I would like to make it so people who work here in the
>office get direct access through the cache directly to the site behind it.
>the content is very dynamic when you're actually working on the site, wo
>using cached content isn't acceptable.
>
>The only thing I saw that might help me here are the always_direct and
>never_direct configuration options, but I've had no success with them.
>
>I tried :
>
>acl office 0.0.0.0/24
>always_direct allow office
>always_direct deny !office
>never_direct allow !office
>never_direct deny office
>
>But this didn't work ... looking at the access.log I'm still seeing
>TCP_HIT:NONE, TCP_IMS_HIT:NONE and so on for my IP, which is in the office
>subnet obviously.
>
>After doing some reading, it looks as if maybe these parameters are only
>useful when sing squid as a proxy, not as an accelerator ...
>
>Did i do something wrong ? What am I missing ? Any suggestions on how I
>might best implement this ?
>
>Thank you,
>
>Jean-François Doyon
>Internet Service Development and Systems Support
>GeoAccess Division
>Canadian Center for Remote Sensing
>Natural Resources Canada
>http://atlas.gc.ca
>Phone: (613) 992-4902
>Fax: (613) 947-2410
>
>
>
Received on Tue May 21 2002 - 09:42:31 MDT