[squid-users] Some questions regarding caching of certain pages.

All,

I saw that Eudora (5.1) was querying our webserver through our squid cache
everytime the program was started [1].

What surprised me that the page is coming from the webserver everytime the program
is started and it looks like being downloaded everytime (I checked it using ngrep
and I can see its coming from the outside).

  Wed Sep 18 18:27:18 2002: schumacher TCP_MISS/200 1489 GET http://www.barrett.com.au/ -
  PARENT_HIT/proxy1.mel.connect.com.au text/html

So I did some tests and loaded some other sites, including:

  www.ugu.com
  www.slug.org.au
  freshmeat.net
  www.barrett.com.au

I loaded each one of them once, closed down netscape and restarted it
and then went "back" to the pages ( I did this multiple time to be sure):

  Wed Sep 18 18:05:53 2002: schumacher TCP_MEM_HIT/200 2619 GET http://www.slug.org.au/slug.css - NONE/- text/css
  Wed Sep 18 18:05:53 2002: schumacher TCP_MEM_HIT/200 7123 GET http://www.slug.org.au/slug-logo.png - NONE/- image/png
  Wed Sep 18 18:05:54 2002: schumacher TCP_MISS/200 8110 GET http://www.slug.org.au/index.html - PARENT_HIT/proxy1.mel.connect.com.au text/html

  Wed Sep 18 18:06:15 2002: schumacher TCP_MISS/200 5643 GET http://www.barrett.com.au/ - PARENT_HIT/proxy1.mel.connect.com.au text/html
  Wed Sep 18 18:06:16 2002: schumacher TCP_MEM_HIT/200 1179 GET http://www.barrett.com.au/mainstyle.css - NONE/- text/plain
  Wed Sep 18 18:06:16 2002: schumacher TCP_MISS/200 7409 GET http://www.barrett.com.au/menu.html - PARENT_HIT/proxy1.mel.connect.com.au text/html
  Wed Sep 18 18:06:17 2002: schumacher TCP_HIT/200 23236 GET http://www.barrett.com.au/scripts/popup.js - NONE/- text/plain
  -- lots more here --

  Wed Sep 18 18:07:29 2002: schumacher TCP_MISS/200 23778 GET http://freshmeat.net/ - PARENT_HIT/proxy1.mel.connect.com.au text/html
  Wed Sep 18 18:07:30 2002: schumacher TCP_MEM_HIT/200 1716 GET http://images.freshmeat.net/button.gif - NONE/- image/gif
  -- lots!!! more --

  Wed Sep 18 18:08:14 2002: schumacher TCP_MEM_HIT/200 2321 GET http://www.ugu.com/ - NONE/- text/html
  Wed Sep 18 18:08:14 2002: schumacher TCP_HIT/200 10877 GET http://www.ugu.com/images/bookbanner.gif - NONE/- image/gif
  Wed Sep 18 18:08:14 2002: schumacher TCP_HIT/200 24128 GET http://www.ugu.com/images/ugu_title_logo.gif - NONE/- image/gif

When I look at the (source of the) html pages they dont have any expire set, yet they
get downloaded from the server every time (except the ugu one), but the *.js
pages come from a MEM_HIT or from the cache (and thankfully all the pics).

It seems the cache is working, yet the html pages get downloaded, is this
done for purpose?

I could understand if the URL has a "cgi" or "?" in it as the cache
is instructed NOT to cache them, but these URL are plain as they can be
and all the pages havent got an expire set.

I even tried it without the upstream parent caches and I had the same
results.

Am I missing something here ???
Any ideas.

jobst

[1] I tried to find out why its doing it but I havent got a clue
    as there is no setting in the preferences.

-- 
C is a write-only language.
             __, Jobst Schmalenbach, jobst@barrett.com.au, Technical Director
   _ _.--'-n_/   Barrett Consulting Group P/L & The Meditation Room P/L      
 -(_)------(_)=  +61 3 9532 7677, POBox 277, Caulfield South, 3162, Australia