On Tue, 10 Feb 2004, Luis Eduardo Cortes wrote:
> I'm using Squid/2.3.STABLE4-hno.CVS
You are aware that this release is not very much tested and not something
I view as suitable for production use are you?
from my old 2.3 patches page
<url:http://devel.squid-cache.org/hno/patch-2.3.html>:
NOTE: I have not tested the Squid-2.3 version as throughtfully as the
Squid-2.2.STABLE5-hno releases (more like not at all), and it is known
that for example async-io will not perform that well compared to
2.2.STABLE5-hno. Most of these patches have been incorporated into the
main Squid distribution since Squid-2.4 and I no longer actively maintain
this patchset.
> By debugging /var/log/messages, I modified iptables rules to accept traffic
> from 127.0.0.1 to 172.19.42.2, and from 172.19.42.2 to 127.0.0.1, and now
> everything it's OK.
>
> My questions are ¿ why suddenly has changed this behavior ? ¿ is this a
> bug ? ¿ can I restore the old behavior ?
Good question.
Maybe there has been some routing related change on your server?
ip ru ls
ip ro ls table 0
Or have you set tcp_outgoing_address? This may maybe make some difference.
Regards
Henrik
Received on Tue Feb 10 2004 - 08:41:54 MST
This archive was generated by hypermail pre-2.1.9 : Mon Mar 01 2004 - 12:00:02 MST