You are probably correct on terms of IE.
Squid is a HTTP/1.0 proxy and as such has to downgrade the reply and
requests to the HTTP version it supports, or else things would break
badly.
httpd_accel_uses_host_header has nothing to do with ETag, only the
Host header.
Note: There is pending patches for ETag support to Squid thanks to
the financial support of another company needing it. Will at least
make Squid use If-None-Match when revalidating it's cache.
Regards
Henrik
On Sunday 19 May 2002 23:12, Warrick FitzGerald wrote:
> Hi All,
>
> Just spent a very long and frustrating day trying to figure out why
> IE is giving me such a hard time with ETags, as was hoping someone
> could shed some light on my painfull day.
>
> - I was initially running Version 2.4.STABLE1 but upgraded to
> Version 2.4.STABLE4.
> - I am running my box as a reverse proxy with a single origin
> server on the back end.
> - client_persistent_connections and server_persistent_connections
> or "on"
>
> I have configed apache with a .htaccess file that looks like this
> (Simply for testing) :
> -----------
> Header append Content-Location: "http://my.test.proxy/"
> Header append Last-Modified "Fri, 17 May 2002 23:56:14 GMT"
> Header append ETag "\"1095d76dfefdc11:897\""
> -----------
>
> 1. When my client (IE 6.0.26) connects directly to the server the
> server it sends a HTTP/1.1 header and recieves a HTTP/1.1 in
> response. 2. The next time the client makes the request it send a
> "If-None-Match" in the header with the ETag it recieved (as per
> HTTP/1.1 spec).
>
> What had me VERY VERY confused today was that when I made this
> request through my squid box, all of a sudden this stopped working
> as expected. After a lot of playing arround I discovered that my
> squid box was making a HTTP/1.0 request back to the origin server
> and even though the origin server responds with a http/1.1 header,
> will still respond to the client with the 1.0 header.
>
> My theory (pls. someone correct me if I'm wrong) is that this
> little problem in turn causes IE not to send the "If-None-Match".
> I'm guessing that this happens because IE sees the response as a
> 1.0 response and decides that the ETag is not part of the spec, so
> just throws it away ?
>
> I then found a param in squid.conf called
> httpd_accel_uses_host_header (Note that I'm pulling straws at this
> point). Tried turning this on and still no joy. I'm am no squid
> expert so I would appreciate if someone could please explain what
> I'm doing wrong.
>
> I have included my config below.
>
> Thanks in advance
> Warrick FitzGerald
>
> [root@WarrickLinux squid]# cat squid.conf | grep -v "^#" | grep
> "\w" hierarchy_stoplist cgi-bin ?
> acl QUERY urlpath_regex cgi-bin \?
> no_cache deny QUERY
> cache_dir ufs /var/spool/squid 100 16 256
> debug_options ALL,9
> acl all src 0.0.0.0/0.0.0.0
> acl manager proto cache_object
> acl localhost src 127.0.0.1/255.255.255.255
> 10.10.52.99/255.255.255.255 acl SSL_ports port 443 563
> acl Safe_ports port 80 # http
> acl Safe_ports port 21 # ftp
> acl Safe_ports port 443 563 # https, snews
> acl Safe_ports port 70 # gopher
> acl Safe_ports port 210 # wais
> acl Safe_ports port 1025-65535 # unregistered ports
> acl Safe_ports port 280 # http-mgmt
> acl Safe_ports port 488 # gss-http
> acl Safe_ports port 591 # filemaker
> acl Safe_ports port 777 # multiling http
> acl CONNECT method CONNECT
> acl PURGE method PURGE
> http_access allow PURGE all
> http_access allow localhost
> http_access allow all
> icp_access allow all
> httpd_accel_single_host on
> httpd_accel_uses_host_header on
> client_persistent_connections on
> server_persistent_connections on
> ie_refresh off
> httpd_accel_with_proxy on
> httpd_accel_uses_host_header on
> http_port 10.10.52.99:80
> httpd_accel_host 10.10.52.129
-- MARA Systems AB, Giving you basic free Squid support Your source of advanced web reverse proxying solutions http://www.marasystems.com/products/Received on Sun May 19 2002 - 17:23:48 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:08:10 MST