New website does not appear on slave server

We are able to add websites to our primary server. When the slave server is designated to receive the site, the site is not added. In fact, the new site for the slave server never gets to the job queue. Both machines are CentOS 5.5 with the same version of ISP Config 3. How can we fix this problem?

 

Results of suggested command

[root@ispconf1 ~]# /usr/local/ispconfig/server/server.sh
26.07.2011-17:41 - DEBUG - Set Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
26.07.2011-17:41 - DEBUG - No Updated records found, starting only the core.
26.07.2011-17:41 - DEBUG - Remove Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
finished.

This is the results of the command on the slave server. Should we have seen the website update here? Should we rerun the website add in the ISPConfig interface and post those results?

 

Rerun suggested command and check log

The output of the "server.sh" command is as follows:
"[root@ispconf1 ~]# /usr/local/ispconfig/server/server.sh
02.08.2011-15:50 - DEBUG - Set Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
02.08.2011-15:50 - DEBUG - No Updated records found, starting only the core.
/usr/bin/fail2ban-client
02.08.2011-15:50 - DEBUG - Remove Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
finished.
You have new mail in /var/spool/mail/root"

Here is an excerp of the slave server log for the same time period:
"02.08.2011-15:40 - DEBUG - Set Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
02.08.2011-15:40 - DEBUG - Found 1 changes, starting update process.
02.08.2011-15:40 - DEBUG - Replicated from master: DELETE FROM web_domain WHERE domain_id = 3
02.08.2011-15:40 - DEBUG - Processed datalog_id 42
02.08.2011-15:40 - DEBUG - Remove Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
02.08.2011-15:41 - DEBUG - Set Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
02.08.2011-15:41 - DEBUG - Found 1 changes, starting update process.
02.08.2011-15:41 - DEBUG - Replicated from master: REPLACE INTO web_domain (`domain_id`,`sys_userid`,`sys_groupid`,`sys_perm_user`,`sys_perm_group`,`sys_perm_other`,`server_id`,`ip_address`,`domain`,`type`,`parent_domain_id`,`vhost_type`,`document_root`,`system_user`,`system_group`,`hd_quota`,`traffic_quota`,`cgi`,`ssi`,`suexec`,`errordocs`,`is_subdomainwww`,`subdomain`,`php`,`ruby`,`redirect_type`,`redirect_path`,`ssl`,`ssl_state`,`ssl_locality`,`ssl_organisation`,`ssl_organisation_unit`,`ssl_country`,`ssl_domain`,`ssl_request`,`ssl_cert`,`ssl_bundle`,`ssl_action`,`stats_password`,`stats_type`,`allow_override`,`apache_directives`,`php_open_basedir`,`custom_php_ini`,`backup_interval`,`backup_copies`,`active`,`traffic_quota_lock`) VALUES ('5','2','2','riud','ru','','3','*','iluvmywebdev.com','vhost','0','name','/var/www/clients/client1/web5','web5','client1','-1','-1','n','n','n','1','1','www','mod','n','','','n','','','','','','','','','','','','webalizer','All','','/var/www/clients/client1/web5/web:/var/www/clients/client1/web5/tmp:/var/www/iluvmywebdev.com/web:/srv/www/iluvmywebdev.com/web:/usr/share/php5:/tmp:/usr/share/phpmyadmin:/etc/phpmyadmin:/var/lib/phpmyadmin','','none','1','y','n')
02.08.2011-15:41 - DEBUG - Processed datalog_id 43
02.08.2011-15:41 - DEBUG - Remove Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
02.08.2011-15:42 - DEBUG - Set Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
02.08.2011-15:42 - DEBUG - No Updated records found, starting only the core.
02.08.2011-15:42 - DEBUG - Remove Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock"

These last 3 lines are constantly replicated.at the end of the file. The web site directories are never installed on the slave server.

 

Follow up on slave server not installing websites.

Any ideas on why this is happening would be greatly appreciated. We are stuck here. Do we need to do more testing?

Thanks.

 

Tutorials used to install the master/slave server config

We used HowToForge Tutorials:
To install the master server we used, " The Perfect Server - CentOS 5.4 i386 ISPConfig 3'. That particular box is a 32 bit machine, so we made sure we used the correct libraries. To install the Multiserver setup for the slave we used, "Installing a Multiserver Setup With Dedicated Web, Email, DNS And MySQL Database Servers on Debian 5.0 with ISPConfig 3". We substituted configuration information for CentOS 5.4 x86 64 ISPConfig 3 for the differences in Debian and CentOS. This is a 64 bit box.

 

The tutorials we used for master/slave

Where should we look next? We made appropiate adjustments as we installed the master server on CentOS 5 while following the second Tutorial for installing master/slave servers in Debian 5.0. Is there something in the difference between Debian and CentOS 5 that may be causing the communication problem? Or is our master server just not filling the slave server queue for some reason? The slave server is recognized by the master in the servers interface.

 

Slave server dB has site record but no directory tree created

There is a record for the new site in the slave server dB that matches the master server dB. But there is no directory tree (/var/www/clients/client1/web5) for the new site on the slave server.

How can we track down this problem?

 

Slave server dB

Should the slave server dB have a directory tree entry or should the master server be writing the directory tree directly to the slave server, bypassing its dB? The slave server dB does not seem to be triggering a cron job to write the tree. Here is a snippet of the cron log:

PHP Warning: mysql_connect(): Unknown MySQL server host 'ispconf.esonicspider.com' (2) in /usr/local/ispconfig/server/lib/classes/db_mysql.inc.php on line 78
PHP Warning: mysql_errno(): supplied argument is not a valid MySQL-Link resource in /usr/local/ispconfig/server/lib/classes/db_mysql.inc.php on line 63
PHP Warning: mysql_error(): supplied argument is not a valid MySQL-Link resource in /usr/local/ispconfig/server/lib/classes/db_mysql.inc.php on line 64
chmod: cannot access `/var/log/ispconfig/httpd/*': No such file or directory
chmod: cannot access `/var/log/ispconfig/httpd/*': No such file or directory

 

Filesystem on slave servers is not touched, you must sync these manually wither with a Clustered FS or something like Rsync...

The apache configs and sys-users are the only thing replicated on a slave server.

 

dB access problem is fixed, still no directory tree

We did see some host file inconsistency on the slave. That was fixed following your instructions. We then renamed the old ispconfig and cron logs so we have fresh logs.
After running another website addition to the slave server the new cron.log is empty. The new ispconfig.log shows no errors and the following update to the slave dB:

"30.08.2011-16:04 - DEBUG - Replicated from master: REPLACE INTO web_domain..."

The log goes on to detail the website we have added but it never gets replicated into the directory tree. We see the dB entries for the new site both in the master dB and in the slave dB. We still don't have the /var/www/clients/... directories.

 

Output of server.sh script on Slave with debugging enabled

Here is the terminal output with no websites waiting to be installed:

"06.09.2011-15:03 - DEBUG - Set Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
06.09.2011-15:03 - DEBUG - No Updated records found, starting only the core.
06.09.2011-15:03 - DEBUG - Remove Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
finished."

Here is the log on the slave server after a website was updated on the master:

"06.09.2011-15:33 - DEBUG - Replicated from master: REPLACE INTO domain (`domain_id`,`sys_userid`,`sys_groupid`,`sys_perm_user`,`sys_perm_group`,`sys_perm_other`,`domain`) VALUES ('1','1','2','riud','ru','','iluvmywebdev.net')
06.09.2011-15:33 - DEBUG - Processed datalog_id 48"

Nothing is written to the slave directory tree.

 

Results of adding website on the Slave Server

I apologise for throwing you off with the previous log copy. I have copied the results of running server.sh after the following info from the current log.

Here is the current ispconfig.log for adding a website to the slave server. There is no directory tree created even though listed in this log:

"13.09.2011-15:36 - DEBUG - Found 1 changes, starting update process.
13.09.2011-15:36 - DEBUG - Replicated from master: REPLACE INTO web_domain (`domain_id`,`sys_userid`,`sys_groupid`,`sys_perm_user`,`sys_perm_group`,`sys_perm_other`,`server_id`,`ip_address`,`domain`,`type`,`parent_domain_id`,`vhost_type`,`document_root`,`system_user`,`system_group`,`hd_quota`,`traffic_quota`,`cgi`,`ssi`,`suexec`,`errordocs`,`is_subdomainwww`,`subdomain`,`php`,`ruby`,`redirect_type`,`redirect_path`,`ssl`,`ssl_state`,`ssl_locality`,`ssl_organisation`,`ssl_organisation_unit`,`ssl_country`,`ssl_domain`,`ssl_request`,`ssl_cert`,`ssl_bundle`,`ssl_action`,`stats_password`,`stats_type`,`allow_override`,`apache_directives`,`php_open_basedir`,`custom_php_ini`,`backup_interval`,`backup_copies`,`active`,`traffic_quota_lock`) VALUES ('11','2','2','riud','ru','','3','*','iluvmywebdev.com','vhost','0','name','/var/www/clients/client1/web11','web11','client1','-1','-1','n','n','n','1','1','www','mod','n','','','n','','','','','','','','','','','','webalizer','All','','/var/www/clients/client1/web11/web:/var/www/clients/client1/web11/tmp:/var/www/iluvmywebdev.com/web:/srv/www/iluvmywebdev.com/web:/usr/share/php5:/tmp:/usr/share/phpmyadmin:/etc/phpmyadmin:/var/lib/phpmyadmin','','none','1','y','n')
13.09.2011-15:36 - DEBUG - Processed datalog_id 58"

The results of running server.sh:

"[root@ispconf1 ~]# /usr/local/ispconfig/server/server.sh
13.09.2011-15:43 - DEBUG - Set Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
13.09.2011-15:43 - DEBUG - No Updated records found, starting only the core.
13.09.2011-15:43 - DEBUG - Remove Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
finished."

 

Directories not created on the slave server

I apologise for throwing you off with the previous log copy. I have copied the results of running server.sh after creating a new website on the master.

The results of running server.sh:

"[root@ispconf1 ~]# /usr/local/ispconfig/server/server.sh
13.09.2011-15:43 - DEBUG - Set Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
13.09.2011-15:43 - DEBUG - No Updated records found, starting only the core.
13.09.2011-15:43 - DEBUG - Remove Lock: /usr/local/ispconfig/server/temp/.ispconfig_lock
finished."

There are still no directories created on the slave server.