[squid-users] TCP_MISS in access log

Hello,

----------------- original message ----------------

I am running BSD/OS 4.2 with the latest stable build of squid. I am trying
to setup a simple proxy/cache using squid. After many configuration
errors I got the client to display a coded web page, and if I put my proxy
and port into netscape and ie I can now see web sites fine. BUT in the
access.log ALL attempts show up at TCP_MISSs and as DIRECT so I am
presuming that although I am using squid as a proxy, I am not using squid
as a cache. Does anyone have any magic to share that is important to get
the cache running? In the ACL section I have listed all of my test
computers, and in the http_access section I gave allow to the test computers.

 I have plenty of harddrive space (16 gig) since I am using a test computer
and no other services are running right now on it..... Also there are NO
hits in the access log so a tcp_miss is not caused by accessing something
that has already been cached.

I am using the following line in squid.conf to indicate the directory for
the cache:

cache_dir ufs /path_to_dir 100 16 256

The directory is owned by nobody and has the permission of rwxr_xr_x
All subdirectories are rwx_rw_rw and also owned by nobody

ACL section I have:
acl all src 0.0.0.0/0.0.0.0
acl myclients src 10.10.10.0/255.255.255.0
acl otherclient src 10.10.11.0/255.255.255.0

HTTP access section:
http_access allow myclients
http_access allow otherclient
http_access allow all

It appears that I am using the proxy part of squid, since all of my request
do fall in the access log, and I can see web pages from other computers
when I put in my host and port in the proxy settings of the browser. But
the log ONLY has tcp_miss and the word DIRECT and the cache directory is
not filling up.

Any tips or hints or help would be highly appreciated

Shera
Received on Fri Jun 22 2001 - 07:52:11 MDT