FW: Loops and tings.

From: Chris Keladis <chrisk@dont-contact.us>
Date: Thu, 20 Aug 1998 13:32:56 +1000

Hi folks,

Just an addendum to my previous email.. The rules I'm now using are:

ipfwadm -I -a accept -S proxy1 -W eth0
ipfwadm -I -a accept -D 0/0 80 -P tcp -r 3128 -W eth0

Still having a lot of these Forwarding loop problems which I cant figure
out.. What possible configuration file combinations can make a cache_peer
that is defined as a sibling, act as a parent ? (or even slightly as a
parent) ?

Chris.

-----Original Message-----
From: Chris Keladis [mailto:chrisk@tpgi.com.au]
Sent: Monday, August 17, 1998 4:04 PM
To: squid-users@ircache.net
Subject: Loops and tings.

Hi folks,

Some problems I've been noticing with our Squid's.. We are running a network
of Squid's in a sibling arrangement with transparent proxy.
I've been noticing this in the log files:
1998/08/17 15:44:47| WARNING: Forwarding loop detected for
'http://spike.scu.edu.au/~gstuart/courses/dp700/' 1998/08/17 15:44:47| -->
1.0 proxy3.tpgi.com.au:3128 (Squid/1.2.beta23), 1.0 proxy2.tpgi.com.au:3128
(Squid/1.2.beta23)
Now, to clarify, at the moment, I am using these ipfwadm rules to make
routing work in the following fashion:
ipfwadm -I -a accept -W lo
ipfwadm -I -a accept -S proxy1 -W eth0
ipfwadm -I -a accept -D 0/0 80 -P tcp -r 3128 -W eth0

My question is, how can I be getting routing loops when the proxy's are in a
Sibling arrangement? It was my understanding sibling relationships ask
neighbouring cache's for an object, and if they return a MISS, and all
subsequent sibling cache's return a miss, then the request will be satisfied
directly by Squid?
This could well be a routing loop on the network somewere, which is an
avenue I am happy to check, but I'm not sure I understand the error message
well enough, to start checking routers, networks, etc.
Any help will be greatly appreciated,

Thanks,

Chris.
Received on Wed Aug 19 1998 - 20:32:34 MDT

This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:41:39 MST