Patrick, I'm sending your message along to the koha and koha-devel mailing lists. You'll likely get better answers more quickly by going directly to these lists. thanks, -pate On Thu, 12 Jun 2003, Singer, Patrick J wrote:
I'm working with Anne Zarinnia's group on setting up a Koha testbed for us to evaluate and test out.
I've been using the 1.9.3 version of Koha because we wanted to get a feel for how the newer stable version (once released) would feel.
I followed the instructions on this page http://koha.org/installation/quick-start.html with some minor changes but it all looks like it is setup correctly.
When I go to the site where I have setup the OPAC the opac-main.pl script crashes an causes an internal server error.
I ran the script from the command line and got the following error message -
HTML::Template->new() : Cannot open included file /default/en/opac-main.tmpl : file not found. at /usr/lib/perl5/site_perl/5.8.0/HTML/Template.pm line 1580
HTML::Template::_init_template('HTML::Template=HASH(0x818ba54)') called at /usr/lib/perl5/site_perl/5.8.0/HTML/Template.pm line 1177
HTML::Template::_init('HTML::Template=HASH(0x818ba54)') called at /usr/lib/perl5/site_perl/5.8.0/HTML/Template.pm line 1071
HTML::Template::new('HTML::Template','filename', '/default/en/opac-main.tmpl','die_on_bad_params',0,'global_vars',1, 'path','ARRAY(0x88218e4)',...) called at /usr/lib/perl5/site_perl/5.8.0/C4/Output.pm line 93
C4::Output::gettemplate('opac-main.tmpl','opac') called at /usr/lib/perl5/site_perl/5.8.0/C4/Auth.pm line 112
C4::Auth::get_template_and_user('HASH(0x806187c)') called at ./opac-main.pl line 12
I don't know if this is due to the includes listed in the /etc/koha.conf file or if it something else. Here is a (close to) complete configuration breakdown of the box.
Software - O/S - RedHat 8.0 Kernel - 2.4.18-14 Apache - 2.0.40-8 mySQL - 3.23.52-3 PERL - 5.8.0-55
Configuration Files -
/etc/koha.conf
database=Koha hostname=localhost user=koha pass=<password> includes=/var/www/koha/htdocs/opac-tmpl
/etc/httpd/conf/httpd.conf | section with opac configuration
# Koha OPAC section ScriptAlias /cgi-bin/koha "/var/www/koha/cgi-bin/" <Directory "/var/www/koha/cgi-bin"> Options +Includes -FollowSymlinks -Indexes AllowOverride None Order allow,deny Allow from all </Directory>
Alias /koha "/var/www/koha/htdocs/" <Directory "/var/www/koha/htdocs"> Options +Includes -FollowSymlinks -Indexes AllowOverride None Order allow,deny Allow from all </Directory>
Directory structure where OPAC located - /var/www/koha/htdocs - location of index.html and directory with opac-tmpl /var/www/koha/htdocs/opac-tmpl - location of opac templates (is this the include directory?) /var/www/koha/cgi-bin - location of opac cgi-scripts
I hope this is enough information to help with the given error.
Thank you for your time and assistance.
-Patrick Singer
Hi Patrick,
Software - O/S - RedHat 8.0 Kernel - 2.4.18-14 Apache - 2.0.40-8
PLEASE : I've a LOT of problems on Mandrake 9.1 with Apache 2.0.43 If you experiment LLLOOOONNNGGG response times in some parts of Koha, tell us. If you don't tell us too. I would like to find if it's a Mandrake, or Apache 2.0.x problem. (LLOONNGG means something like 5-7 minuts : the httpd script timeout length it seems)
mySQL - 3.23.52-3 PERL - 5.8.0-55
Configuration Files -
/etc/koha.conf
database=Koha hostname=localhost user=koha pass=<password> includes=/var/www/koha/htdocs/opac-tmpl
those lines are missing in koha.conf : intrahtdocs=/home/paul/koha.dev/koha/koha-tmpl/intranet-tmpl opachtdocs=/home/paul/koha.dev/koha/koha-tmpl/opac-tmpl
-- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)
On Wed, 18 Jun 2003, paul POULAIN wrote:
If you experiment LLLOOOONNNGGG response times in some parts of Koha, tell us. If you don't tell us too. I would like to find if it's a Mandrake, or Apache 2.0.x problem. (LLOONNGG means something like 5-7 minuts : the httpd script timeout length it seems)
One of the things I found in moving to Apache 2.0 was that if my scripts produce more than a certain volume of error logging per invocation things locked up for extended periods of time "inexplicably". This was running Perl scripts in a CGI fashion on RH8. I'm experimenting with the same Perl code in mod_perl 1.99 and I haven't had the same problem crop up, but I haven't pushed the limits with it. Nothing shows up in the 2.0 ChangeNotes for the last 10 minor releases of apache that seems to reference anything like this. I haven't been able to find anything on google indicating that anyone else has experienced this either. If you happen to find you're having a similar problem and reducing the amount of info going to the error log "fixes" your problem please let me know and I'll dig a little harder in a few more places. -- </chris> The death of democracy is not likely to be an assassination from ambush. It will be a slow extinction from apathy, indifference, and undernourishment. -Robert Maynard Hutchins, educator (1899-1977)
Christopher Hicks wrote:
On Wed, 18 Jun 2003, paul POULAIN wrote:
If you experiment LLLOOOONNNGGG response times in some parts of Koha, tell us. If you don't tell us too. I would like to find if it's a Mandrake, or Apache 2.0.x problem. (LLOONNGG means something like 5-7 minuts : the httpd script timeout length it seems)
One of the things I found in moving to Apache 2.0 was that if my scripts produce more than a certain volume of error logging per invocation things locked up for extended periods of time "inexplicably". This was running Perl scripts in a CGI fashion on RH8. I'm experimenting with the same Perl code in mod_perl 1.99 and I haven't had the same problem crop up, but I haven't pushed the limits with it.
Nothing shows up in the 2.0 ChangeNotes for the last 10 minor releases of apache that seems to reference anything like this. I haven't been able to find anything on google indicating that anyone else has experienced this either. If you happen to find you're having a similar problem and reducing the amount of info going to the error log "fixes" your problem please let me know and I'll dig a little harder in a few more places.
That's almost the same problem for me... A sample : if I call in a script a MARC::Record.pm function with wrong or invalid parameters, the pm is supposed to return a warning, and die. With apache 2.0, it seems to be running and running until httpd timeout. Then, the log is filled with hundreds of warnings (the one reported by .pm, repeated). It seems that the process forks, and forks, and forks forever, until apache stops it. I tried to ask on some mailing lists, but could not get any answer. Feel free to dig yourself an report any success in your search. -- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)
On Thu, 19 Jun 2003, paul POULAIN wrote:
It seems that the process forks, and forks, and forks forever, until apache stops it.
Are you sure it's forking? Can you try to comment out the bulk of the warnings being sent to the logs and see if it gets any better? Reducing the volume of warnings being generated is the only workaround I've found currently. -- </chris> The death of democracy is not likely to be an assassination from ambush. It will be a slow extinction from apathy, indifference, and undernourishment. -Robert Maynard Hutchins, educator (1899-1977)
Christopher Hicks wrote:
On Thu, 19 Jun 2003, paul POULAIN wrote:
It seems that the process forks, and forks, and forks forever, until apache stops it.
Are you sure it's forking? Can you try to comment out the bulk of the warnings being sent to the logs and see if it gets any better? Reducing the volume of warnings being generated is the only workaround I've found currently.
Not sure it's a unix fork. Sure it produces hundreds of lines, where apache 1.3.x produces 1 or 2. I've uninstalled Mandrake 9.1 so I can't test anymore. -- Paul POULAIN Consultant indépendant en logiciels libres responsable francophone de koha (SIGB libre http://www.koha-fr.org)
participants (3)
-
Christopher Hicks -
Pat Eyler -
paul POULAIN