On 23/07/11 04:24, Ritter, Nicholas wrote:
I should add one important point. When the error occurs, it is most
often not affecting the entire site nor transaction. This is to say that
I can visit a site, get content, and then at some point fill out a form
on the site, which then generates the 502. I don't want anyone to assume
that the 502 is being generated because of an obvious path connectivity
error where the site being surfed was down all along.
I should also not that I am not running any unique refresh patterns in
the squid.conf.
-----Original Message-----
From: Ritter, Nicholas [mailto:nicholas.rit...@americantv.com]
Sent: Friday, July 22, 2011 11:16 AM
To: squid-users@squid-cache.org
Subject: [squid-users] squid 3.1 (w/ TPROXY/WCCP) and increased 502 and
206 codes
I am doing extended testing of a CentOS v6 TPROXY/SQUID3/WCCP setup and
I noticing higher than usual TCP_MISS/502 codes. I am also seeing some
206 codes, but it is the 502s that are much higher than normal. I think
it is transport related inside the TPROXY/SQUID side of things but I am
not sure.
I am seeing the 502 codes on both gets and posts. Can anyone provide
more insight on this condition and what/where I should start
troubleshooting?
With the message presented in that 502 error page. 502 is sent on
several outbound connection problems from TCP connect through to reply
parsing.
I am running the stock CentOS v6 kernel (2.6.32-71.29.1) and Squid
3.1.10 as package by RHEL 6 (specifically a RHEL 6 rebuilt source rpm of
squid-3.1.10-1.el6.
Should update to the more recent release of squid 3.1 as a starting
point?
Always a good choice to know if its been fixed. Though I don't recall
anything major having changed since .10 regarding connectivity.
Amos
--
Please be using
Current Stable Squid 2.7.STABLE9 or 3.1.14
Beta testers wanted for 3.2.0.9
