Re: [squid-users] Causes for RELEASE Action in store.log

From: Norbert Hoeller <nhoeller_at_sinet.ca>
Date: Thu, 26 Nov 2009 11:08:58 -0500

> Your squid.conf default settings might have "acl QUERY url_regex cgi-bin
> \?", "cache deny QUERY", or the nicer "refresh_pattern -i (/cgi-bin/|\?)
> 0 0% 0" to tell squid explicitly what requests to consider dynamic. But
> the defaults are not involving .shtml

I did not find any "cache deny QUERY" in squid.conf.  I have the default refresh_pattern settings:

  #Suggested default:
  refresh_pattern ^ftp:           1440    20%     10080
  refresh_pattern ^gopher:        1440    0%      1440
  refresh_pattern (cgi-bin|\?)    0       0%      0
  refresh_pattern .               0       20%     4320

> The website itself is being nasty and telling Squid that those whole
> pages change between browsers. "Vary: User-Agent" which can make Squid
> drop objects when a different browser from the one which caused storage
> earlier requests it.
> NP2: Internet Explorer is particularly bad with this since each PC users
> installed Ad-Ons, their versions, the patches applied to IE in that box
> and that users account, and the exact build time of the IE binary are
> all stored in the User-Agent text.

I do not see anything unusual in the page source.  Here is that metadata:

 <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">

I just did a test: loaded the page in Firefox, closed the tab, opened a new tab and loaded the page in Firefox again. squid reported a TCP_MISS on the shtml page and a TCP_IMS_HIT on a CSS file. It appears Firefox cached all the other objects on the page except for these two. Strangely, store.log shows a RELEASE for both files. I will do some more testing and check out any squid debug options.
      Thanks, Norbert
Received on Thu Nov 26 2009 - 16:09:06 MST

This archive was generated by hypermail 2.2.0 : Thu Nov 26 2009 - 12:00:03 MST