To answer the earlier suggestions; 

* Yes, logged in as root to run datagard.
* All Midgard directories, files and permissions are _identical_ on "happy" and
"sad". To make sure I've got everything, this includes everything under the
following:

Everything under:
 /etc/midgard/
 /var/cache/midgard/
 /var/lib/midgard/
 /usr/share/midgard/
the 4 midgard .so files under:
 /usr/lib/apache2/modules
the 2 libmidgard files under:
 /usr/lib
the 5 midgard files under:
 /usr/bin
and /usr/lib/php4/20020429/midgard.so

/etc/apache2/apache2.conf and /etc/php4/apache2/php.ini are identical on both
machines. When I checked out the latter, however, I did notice that the php.ini
file in sad did _not_ include the line extension=mysql.so, though this was in
"happy". I added it, dumped and reinstalled the Midgard database using datagard,
restarted mysql and apache2 and... same error.

The database and virtual hosts were set up with datagard. Usernames and
passwords are the same on both.

The one test that sticks out comes from running midgard-pageresolve, which as I
noted produces quite different results on "happy" and "sad". I couldn't find any
mention of midgard-pageresolve in the docs or on www.midgard-project.org, so
don't know how to interpret the output (see
http://article.gmane.org/gmane.comp.web.midgard.user/6611 for my results)

It does seem odd, however, that the _working_ server reports "RESULT: No host
record match", but the _non-working_ server reports "Page found". 

I'm trying to isolate the problem, but at am a loss. Does anything here hint as
to what might be going wrong? I just can't think what else could be causing this
behaviour.

B.


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to