3.1.0.16 errors/ca/ERR_CACHE_ACCESS_DENIED empty

2010-02-07 Thread Henrik Nordström
Something seems to have gone wrong with error page generation in the
3.1.0.16 tarball. errors/ca/ERR_CACHE_ACCESS_DENIED is empty.

Regards
Henrik



Re: 3.1.0.16 tarball mismatch?

2010-02-07 Thread Henrik Nordström
sön 2010-02-07 klockan 15:43 +1300 skrev Amos Jeffries:

 Yes, there was for a few hours a bundle that did not build.
 The signature got as far as HTTP east/west before I erased everything 
 and started again.

And I picked up the first when building 3.1.0.16 for Fedora, which
caused some major confusion.

Regards
Henrik



WCCPv1 still broken in squid-3.1

2010-02-07 Thread Graham Keeling
Hello,

I started a thread over here with the initial problems that I had with
squid-3.1.0.15:
http://www.mail-archive.com/squid-us...@squid-cache.org/msg69906.html

That thread ends with me setting the router port in src/wccp.cc and I say
that WCCPv1 was working for me.

Well, I was mistaken, as it still didn't work (it turned out that I hadn't
actually stopped a previously running squid-3.0 process).

However, I have another patch that really *does* make it work for me (in
addition to the original port patch).


It seemed to me that wccpAssignBuckets() never gets called, and that this
was because wccpLowestIP() never returned non-zero.
And this in turn seemed to be because the ip_addr and local_ip variables in
wccpLowestIP() were being compared, but 'IpAddress local_ip' was getting
initialised from a 'struct addrinfo *';

local_ip = *local;

I changed it to the following, and suddenly my WCCPv1 browsing works:

local_ip.InitAddrInfo(local);

I don't know whether this is the 'correct' thing to do or not, as I don't
understand squid's 'class IpAddress'. Perhaps there needs to be an equivalent
FreeAddrInfo somewhere to stop it eating memory.

But I do know that I can now use WCCPv1, whereas before it didn't work at all.

Patch attached.
Index: wccp.cc
===
RCS file: /cvs/netpilot/GPL/squid-3.1.0.15/WORK/src/wccp.cc,v
retrieving revision 1.2
diff -u -r1.2 wccp.cc
--- wccp.cc	3 Feb 2010 14:36:30 -	1.2
+++ wccp.cc	3 Feb 2010 16:46:38 -
@@ -177,7 +177,7 @@
 if (getsockname(theWccpConnection, local-ai_addr, local-ai_addrlen))
 fatal(Unable to getsockname on WCCP out socket);
 
-local_ip = *local;
+local_ip.InitAddrInfo(local);
 
 Config.Wccp.address.FreeAddrInfo(local);
 }


Delay pool implementation

2010-02-07 Thread Wathsala Vithanage
Hi,

I'm working on a project aimed at developing dynamic/intelligent
bandwidth controlling
techniques as a part of my Masters work.
I'm developing a system that intelligently decides the maximum amount
of bandwidth
allowed for a given connection, I would like to extend Squid delay
pools as a POC.
Are there any resources where I can learn about Squid delay pool implementation
other than the source code.

Regards.


Re: [PATCH] Plain Surrogate/1.0 support

2010-02-07 Thread Robert Collins
On Sun, 2010-02-07 at 00:52 +1300, Amos Jeffries wrote:
 According to the W3C documentation the Surrogate/1.0 capabilities and 
 the Surrogate-Control: header are distinct from the ESI capabilities.
 
 This patch makes Squid always advertise and perform the Surrogate/1.0 
 capabilities for reverse-proxy requests.
 
 Full ESI support is no longer required to use the bare-bones 
 Surrogate-Control: capabilities.

A quick check though - is it still only enabled for accel ports? (It
shouldn't be enabled for forward proxying).

-Rob


signature.asc
Description: This is a digitally signed message part