Error 500 is an error of web server, check web server configuration,
may be there is a limit for upload via POST.

Lorenzo

On 4/4/07, Mike Ellery <mikee(at)s2technologies.com> wrote:[...]

Hi,

We found it easier to simply re-create horde/imp in house than to try 
and run horde/imp in our cluster(webmail is not rocket science). We did 
not spend a lot time trying to make imp work .....

Jimmy


Charlie Schluting wrote:[...]

Dave Steinberg schrieb:[...]
They will be found by the brave people who try the experimental branches
because they need the features *right now* :).


thomas

On Wed, 2007-04-04 at 11:00 -0700, Stan Cooper wrote:[...]

All requests to a specific URL can be redirected to another URL (which
may correspond to a script or anything else).[...]

Stan, I can't see how that is ever going to run, pound can't run cgi
(perl, python et al) itself?

I guess what you actually want is to pass the request to Apache, running
mod_python so that the ldap-request script can run with the request
parameters?

 John Snowdon - IT Support Specialist
-==========================================-
 School of Medical Education Development 
 Faculty of Medical Sciences Computing
 University of Newcastle

 Email : j.p.snowdon(at)ncl.ac.uk

[...]

On Tue, 2007-04-24 at 16:36 -0400, jerome.doucerain(at)bell.ca wrote:[...]

Quite a few problems were fixed since 2.1 - I suggest you upgrade to
2.3. Please let us know if you still see the problem after the upgrade.[...]

Here's a concrete example:

http://www.arch.columbia.edu/DDL/cad/A4535/F00cad/textures/images/Concrete.jpg

;-)

Glenn

Víg Pál wrote:[...]

On Wed, 2007-04-25 at 19:12 +0200, Víg Pál wrote:[...]

You need to use the correct regular expression for this - something like
"([^.].*)|(.[^.].*)" to match anything that doesn't start with "..".

It would probably easier to match ".. whatever" first in the service
that actually answers to it (possibly without a back-end if you want it
permanently denied) and then match everything else:

Service
	URL "\.\./.*"
# optional Backend if you want to actually do something with these
# requests
#	Backend
#		...
#	End
End

Service
# Requests get here only if they don't start with ../
	Backend
		...
	End
End
[...]

That is not a Pound function. Some web servers automatically issue a
redirect in response to requests for a directory. Check on what redirect
is issued, and if Pound rewrites it correctly. BTW, "works" and "does
not work" is not very descriptive.

If you want Pound to issue these redirects use a Redirect, matching
exactly your URL (enclosed in ^ and $). This is bad design and not
recommended.[...]

On Thu, 2007-04-26 at 13:33 +0200, RedShift wrote:[...]

This is probably hilarious to you, but I must ask you to keep this on a
professional level. The OP asked a legitimate question, so reply if you
have an answer.[...]

Robert Segall wrote:[...]


We have a situation where we needed to negate the URL as well.  Instead
of trying to figure out a negative regular expression, we ended up doing
the two service method you describe here as well.  Our reasoning was we
have certain URL paths that are testing paths and should only be
accessible to internal users, who are behind the firewall anyway.  So
our first service definition matches valid external customers and allows
them to pass through fine.  The second service is the bit bucket and the
requests are logged.

Kenneth

On Mon, 2007-04-30 at 11:30 +0100, Rachel wrote:[...]

Do your back-ends actually issue a header called X-sess? Is the value in
X-sess unique per session (regardless of back-end)? If not the above
configuration won't help much.[...]

On 4/30/07, Robert Segall <roseg(at)apsis.ch> wrote:[...]


I have setup my application to pass that header with the unique session ID:


<?php

session_start();
$session = session_id();

header('X-sess: ' . $session);

?>

<p>Welcome to my load balancer test site!</p>

<?php


echo $_SERVER['SERVER_NAME'];


echo ' - ' . $session;

?>

Thanks



Rachel

On 4/30/07, Rachel <didlix(at)gmail.com> wrote:
[...]


Is this not a suitable method of doing this? Should the X header be sent by
the backend server and not PHP?

I can of course use cookie's instead.

Thanks

Rachel


[...]

On Mon, 2007-04-30 at 13:56 +0100, Rachel wrote:[...]

I am no PHP expert, but I seem to recall that session_start() sets a
cookie. Do you see this cookie (and the X-sess header) in the browser?
How about between Pound and the back-ends?[...]

The php_session temporary data are stored as files on each server. I think
that cookies would be the best way to manage this.

Tyler Collard

On 4/30/07, Robert Segall <roseg(at)apsis.ch> wrote:[...]

On 4/30/07, Tyler Collard <tcollard(at)gmail.com> wrote:[...]


So using the php header function wont work (which seems likely as, pound is
not keeping affinity with one server even though the header is being sent by
php)

>[...]


You are correct that session_start() set's a cookie,
The reason I have opted for header is in case the user has cookies disabled.

The header and cookie is being seen by the web browser as LiveHTTPHeaders
shows:

HTTP/1.x 200 OK
Date: Mon, 30 Apr 2007 10:54:00 GMT
Server: Apache/1.3.37 (Unix) PHP/5.2.1
X-Powered-By: PHP/5.2.1
Expires: Thu, 19 Nov 1981 08:52:00 GMT
Cache-Control: no-store, no-cache, must-revalidate, post-check=0,
pre-check=0
Pragma: no-cache
X-sess: 211ccb1d6235d98a9c40ae659c8ea4ee



Thanks

Rachel

On Mon, 2007-04-30 at 15:21 +0100, Rachel wrote:[...]

That would be a Pound bug, which is surprising as quite a few people use
this. Could you possibly use poundctl to look at the sessions?[...]

Hi,
 
Rachel wrote:[...]

It seems clear that the PHP "response header" is being seen by the
browser, but is the browser sending it to Pound on every request as a
"request header", so that Pound knows what back-end server to hit?

Just a thought,

 Curro

This email message is intended only for the use of the named recipient.
Information contained in this email message and its attachments may be
privileged, confidential and protected from disclosure. If you are not the
intended recipient, please do not read, copy, use or disclose this
communication to others. Also please notify the sender by replying to this
message and then delete it from your system.

# /opt/pound-2.3/sbin/poundctl -c /opt/pound-2.3/socket
  0. http Listener 192.168.1.68:80 a
    0. Service active (10)
      0. Backend PF_INET 192.168.2.70:80 active (5 0.000 sec) alive
      1. Backend PF_INET 192.168.2.71:80 active (5 0.000 sec) alive
      0. Session 211ccb1d6235d98a9c40ae659c8ea4ee -> 0
 -1. Global services

Hitting F5 repeatedly will randomly give me one of the 2 back end servers
but the session is always the same.



On 4/30/07, Robert Segall <roseg(at)apsis.ch> wrote:[...]

>[...]


I don't think it would which would explain it :o)