Re: no_cache for local www

From: Stuart Young <sgy@dont-contact.us>
Date: Mon, 11 Jan 1999 13:13:24 +1100

At 19:53 10/01/99 +0100, Emmanuel Keller wrote:
>always a computer on the local network access the intranet www server,
>the name of the computer which access the www server is cache.

Sure you haven't got some process fetching files from the www server using
wget or something?

I've also had a few people tell me that they can use BNC (an irc
redirector) to allow connections to any machine/port from a machine running
squid, wether BNC is running on the squid box, or not (and therefore
requests are seen as coming from the cache), overriding any rules you have
in the process. If someone can verify this, please post here.

Of course, you could just block cache from your www server... that'd be
simple enough.

Otherwise I reccomend setting up a machine that has some sort of
tcpdump/libpcap program on it (eg: ethereal - Check www.freshmeat.org for
details on where to obtain it), and log all data to/from the cache and www
server for a day or two. This gives you 2 important things.. 1> a log of
that exactly is going on, and therefore an understanding of why/how/etc.
2> a record of data communication to send to the squid maintainers if this
is a legitimate bug.

Stuart Young - sgy@amc.com.au - cefiar@amarok.glasswings.com.au
(aka Cefiar) - http://amarok.glasswing.com.au/

[All opinions expressed in the above message are my]
[own and not necessarily the views of my employer..]
Received on Sun Jan 10 1999 - 18:56:42 MST

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:43:57 MST