Squidguard Error

[jaybee]

I am using squidguard-devel and squid3 (although i also have the same problem on squid and squidguard beta).  When I go to youtube.com, the top banner ad is replaced with:

ERROR
The requested URL could not be retrieved

The following error was encountered while trying to retrieve the URL: http://172.16.1.1:88/sgerror.php?

Zero Sized Reply

Squid did not receive any data for this request.

Your cache administrator is admin@localhost.

The strange thing is that 172.16.1.1:88 is the address of the admin gui.  If I go to admin access page in the gui and blank out the TCP port field then I don't get the error.  If I change the port to anything but blank then the error above comes up on youtube.com only (that I know of) but with the 88 replaced by what ever port I changed the admin GUI to.  What is the world does having a port specified for the admin gui have to do with this squid error?

[jaybee]

UPDATE:  If I wait several minutes once going to youtube.com.  The blank space at the top where the ad normally is will give me the following error if I have the TCP port field blank for the admin gui port. If I specify a port for the admin gui then I get the error in the previous message immediately.

ERROR
The requested URL could not be retrieved

The following error was encountered while trying to retrieve the URL: http://172.16.1.1/sgerror.php?

Connection to 172.16.1.1 failed.

The system returned: (60) Operation timed out

The remote host or network may be down. Please try the request again.

Any ideas?

[Legion]

I'm messing around with squidGuard and came across the same error. It's coming from these lines in squidguard_configurator.inc:

# -------------------------------------------------------------------------------------------------
# sg_redirector_base_url
#   $url - url where redirect to
#   $use_internal - ignore 'Redirect mode' option, use internal (for rebuild config, for example)
#
# -------------------------------------------------------------------------------------------------
function sg_redirector_base_url($rdr_info, $redirect_mode)
{
    global $squidguard_config;
    $rdr_path = '';

    # gui port, ip & proto
    $guiip = (!empty($squidguard_config[F_CURRENT_LAN_IP])) ? $squidguard_config[F_CURRENT_LAN_IP] : '127.0.0.1';
    $guiport = (!empty($squidguard_config[F_CURRENT_GUI_PORT])) ? $squidguard_config[F_CURRENT_GUI_PORT] : '80';
    $rdr_path = "http://$guiip:$guiport" . REDIRECT_BASE_URL;
    ...

Since most of us are probably using https for the webgui, this code won't find sgerror.php. I happen to be running a lighttpd server on port 80 as well, but this code won't find that either because it will use my https webgui port.

I don't know php or how much about pfSense state the code can determine at this level. Is it possible to know if the webgui is in https mode? If so, $rdr_path should reflect that.

[Legion]

Actually, I tried editing my squidguard_configurator.inc code to use https and got certificate errors. So in my case it works best if I leave the protocol the same but hardcode $guiport to '80'. But this relies on my having my second lighttpd running on 80 and won't work for most users. However, it now presents the error messages I expect, using:

# -------------------------------------------------------------------------------------------------
# sg_redirector_base_url
#   $url - url where redirect to
#   $use_internal - ignore 'Redirect mode' option, use internal (for rebuild config, for example)
#
# -------------------------------------------------------------------------------------------------
function sg_redirector_base_url($rdr_info, $redirect_mode)
{
    global $squidguard_config;
    $rdr_path = '';

    # gui port, ip & proto
    $guiip = (!empty($squidguard_config[F_CURRENT_LAN_IP])) ? $squidguard_config[F_CURRENT_LAN_IP] : '127.0.0.1';
    # $guiport = (!empty($squidguard_config[F_CURRENT_GUI_PORT])) ? $squidguard_config[F_CURRENT_GUI_PORT] : '80';
    $guiport = '80';
    $rdr_path = "http://$guiip:$guiport" . REDIRECT_BASE_URL;
    ...

[jolebole]

I am surprised that there is no  resolution for this problem yet. Some recommend use http for the admin gui and the problem goes away or use an external error page, which might be a second best possible option.

[doktornotor]

Why are you reviving zombie threads? :(

[jolebole]

Well l, I searched for a solution. Found 10 threads with the same problem and no solution. Didn't even noticed its a buried thread.

[Snailkhan]

me also facing similar error.
I get this error for sites that are blocked.
@OP
were you able to find a solution for it ?

[exa_gon]

Someone have find a solution ? I 've the same error.

[rharder]

Zombie thread or not, this long-standing problem without a solution is crazy.

With squid == websites broken

Without squid == websites work

Apparently browsers don't mind having dropped connections or whatever else is going on.  Why does squid cause a greater failure than apparently the source error?

-Rob