Emiliano wrote:
> OK, so mod_midgard is definately causing this. The requests you were
> placing,
> was mod_midgard supposed to handle these? Because the earlier log showed
> that mod_midgard thought it didn't. I'll create a patch that does more
> extensive logging and we'll find the problem.
OK, I've been goign trhough the code that from the looks of it is the
cause
of these problems. I can't find the troublespot.
> [Fri Jan 19 20:21:07 2001] [warn] pid file /var/run/httpd.pid overwritten --
> Unclean shutdown of previous Apache run?
> [Fri Jan 19 20:21:07 2001] [info] created shared memory segment #4609
> [Fri Jan 19 20:21:13 2001] [debug] mod_midgard.c(1178): [client
> 213.25.250.81] midgard_translate_handler(russian,
> http://wanpipe.pasjo.net.pl:8101/admin)
> [Fri Jan 19 20:21:13 2001] [debug] /usr/include/mgd_apache.h(135): [client
> 213.25.250.81] Midgard: select_database opened midgard
> [Fri Jan 19 20:21:13 2001] [debug] /usr/include/mgd_apache.h(123): [client
> 213.25.250.81] Midgard: select_database kept current (midgard)
This looks OK, mod_midgard is handling the request but since the URL is
not
absolute (ends in /admin instead of /admin/) it is sending a redirect.
Since
the log continues without interruption I assume apache didn't segfault
at
this point. However, I did try to contact this server myself and while
it
does answer my request I get no data sent back. Could you try the
following:
$ telnet wanpipe.pasjo.net.pl 8101
+Trying 213.77.2.81...
+Connected to wanpipe.pasjo.net.pl.
+Escape character is '^]'.
GET /admin HTTP/1.1<return>
Host: wanpipe.pasjo.net.pl<return>
<return>
to see what you get back in both telnet and the error log, with debug
enabled please. The lines starting with '+' is output, you don't type
those. The second <return> is necesary.
> [Fri Jan 19 20:21:13 2001] [debug] mod_midgard.c(1178): [client
> 213.25.250.81] midgard_translate_handler(russian,
> http://wanpipe.pasjo.net.pl:8101/errordocs/401.shtml)
> [Fri Jan 19 20:21:13 2001] [debug] mod_midgard.c(1221): [client
> 213.25.250.81] Midgard: host record for wanpipe.pasjo.net.pl:8101 not found
There are a few things I still don't understand here.
First, I would have expected a request to
http://wanpipe.pasjo.net.pl:8101/admin/ here.
Second, I wonder what caused the browser to fetch
http://wanpipe.pasjo.net.pl:8101/errordocs/401.shtml.
Third, immediately after Midgard emits this last message it passes
control
back to Apache and isn't reactivated for this request so I just can't
even
find a possible troublespot that would cause the segfault. Immediately
after
this message, midgard in its entirity is out of the loop.
If anyone can shed some light on this, maybe seen parts or all of this
happen before, let me know.
Emile
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
