I'm using 3.5 with transparent server first bumping in ~100 deployments so
far, it works just fine, excluding with SNI and everything.
On 12/07/2015 10:58 am, Yuri Voinov yvoi...@gmail.com wrote:
Man,
3.5.x don't work with server-first. It must be for backward compatibility
- but don't be.
()
Hope that helps,
Nathan.
On 30 May 2015 at 01:14, James Lay j...@slave-tothe-box.net wrote:
On 2015-05-29 08:57 AM, Nathan Hoad wrote:
Yes, I have it working on about a dozen deployments so far, using an
external ACL to make bumping decisions based on the SNI server name
and a few other
Yes, I have it working on about a dozen deployments so far, using an
external ACL to make bumping decisions based on the SNI server name and a
few other things. No complaints from me, it Just Works.
On 29/05/2015 5:50 pm, sp_ ap...@yandex.ru wrote:
Hello,
does anyone have the working squid 3.5
script example you provided to work for me.
Does the new server_name acl change how this can be done? Would you be able
to provide a new example for me to try based on this new acl in squid 3.5.4?
Any help is greatly appreciated.
Stan
On Sun, Apr 12, 2015 at 7:25 PM, Nathan Hoad nat
These are fixed in 3.5.4.
Nathan.
On 5 May 2015 at 08:59, Prashanth Prabhu prashanth.pra...@gmail.com wrote:
Hi folks,
I am seeing a bunch of Squid crashes after moving recently to 3.5.1.
The stack trace for the various crashes is below, along with info I
have been able to gather for them.
Hi Fred,
I believe you're experiencing bug 3329:
http://bugs.squid-cache.org/show_bug.cgi?id=3329
Please try the patch that I have on there and see if that helps your issue.
Thank you,
Nathan.
On 17 April 2015 at 17:06, Stakres vdoc...@neuf.fr wrote:
Hi All,
Is anyone with an trick
As I mentioned earlier, this hasn't worked the because the debug
symbols you're running gdb against don't quite match the binary in
which the crash occurred. I would check the version of the debug
symbols you're running gdb against versus the version of the squid
binary that was installed. These
What version of Squid are you running, and in what environment?
Based on the fact that the output of the gdb backtrace doesn't show
function parameters (debug symbols don't quite match up to the
binary), this may not work, but if you could run these commands in gdb
and show the output, that would
Hello,
I believe you're experiencing bug 3329:
http://bugs.squid-cache.org/show_bug.cgi?id=3329
Please try the patch that I have on there and see if that helps your issue.
Thank you,
Nathan.
On 15 April 2015 at 07:16, HackXBack hack.b...@hotmail.com wrote:
[New LWP 4677]
[New LWP 4704]
Hi Stan,
For peek and splice, you need to decide based on the SNI name, not the
domain name, which for 3.5 means you need to use an external ACL
helper that processes %ssl::sni. For 4.0 there will be a server_name
ACL you can use instead.
On top of that, you also need to make sure this external
is actually used to
identify the sites to not bump?
Is there a good write-up somewhere of how to create these ACL helpers and
how to use them?
On Sun, Apr 12, 2015 at 8:25 PM, Nathan Hoad nat...@getoffmalawn.com
wrote:
Hi Stan,
For peek and splice, you need to decide based on the SNI name
Hello,
Do you have a core dump? If so, if you could get a backtrace as
described here http://wiki.squid-cache.org/SquidFaq/BugReporting and
post it here?
Thank you,
Nathan.
On 2 April 2015 at 03:45, Stakres vdoc...@neuf.fr wrote:
Hi All,
Strange problem during surf, squid 3.5.3, 64bits,
with the external ACL helper.
Cheers,
Nathan.
On 9 March 2015 at 16:06, Amos Jeffries squ...@treenet.co.nz wrote:
On 9/03/2015 5:52 p.m., Nathan Hoad wrote:
Hi folks,
I'm playing with 3.5.2 and Peek-n-Splice, I was wondering if it's
actually possible to exclude requests based on the SNI host and have
13 matches
Mail list logo
