Ok good. I think there is a misbehavior behind that. I will continue to check on my side why the forgeupgrade didn't run properly.

That solved it. Thanks a lot for your help.

As you are not in production, you can try to remove the fact that last forgeupgrade were run and try to relaunch the forgeupgrade.

Please do DELETE FROM forge_upgrade_bucket WHERE ID > 126; and redo the /usr/lib/forgeupgrade/bin/forgeupgrade --config=/etc/codendi/forgeupgrade/config.ini update to see what happen

The value of $sys_dbuser is codendiadm.
Yes, I have a codendiadm user in the database. Here is the output of SHOW GRANT when I use the codendiadm user in MySQL:
GRANT USAGE ON *.* TO 'codendiadm'@'localhost' IDENTIFIED BY PASSWORD '5c449122063c3059'
GRANT ALL PRIVILEGES ON `tuleap`.* TO 'codendiadm'@'localhost'
GRANT ALL PRIVILEGES ON `plugin_mediawiki_%`.* TO 'codendiadm'@'localhost'
GRANT ALL PRIVILEGES ON `cx_%`.* TO 'codendiadm'@'localhost' WITH GRANT OPTION

The import of data does not imply only the DB, but also the filesystem (uploaded files and so on).

You don't have a codendiadm user in your DB ? What is the value of $sys_dbuser in /etc/tuleap/conf/database.inc ?

1. I have used the root user

2. Output is:
   -rw-r--r--. 1 codendiadm codendiadm 17847 Aug 15 12:52 /etc/tuleap/conf/local.inc
   -rw-r--r--. 1 codendiadm codendiadm 17847 Aug 15 12:52 /etc/codendi/conf/local.inc

3. SHOW GRANTS returns:
   GRANT ALL PRIVILEGES ON *.* TO 'root'@'localhost' IDENTIFIED BY PASSWORD 'xxxx' WITH GRANT OPTION
   And the SELECT returned: 0

BTW: this tuleap server is not running in production yet. So I could reset the database content if necessary (and possible). I only would like to avoid a complete reinstall. Or is ist possible to export the database content on one machine and then import it on another? I have got a running tuleap instance on another machine and I am thinking about transferring all stored data to the new machine.
 

Ok so I've got some other questions:

1. Which unix user runs the forgeupgrade script ?
2. What is the output of: ls -al /etc/tuleap/conf/local.inc AND ls -al /etc/codendi/conf/local.inc ?
3. On MySQL, please run these two queries: SHOW GRANTS;  and SELECT count(*) FROM forge_upgrade_log;

This is the expected behaviour as your forgeupgrade is not not properly runned, this table must be missing.

Two more questions :

1. What is the content of /etc/tuleap/forgeupgrade/config.ini ?
2. What is the result of this MySQL query: SELECT * FROM forge_upgrade_log WHERE bucket_id > 126; ?

Ok.

Can you provide me the output of these two SQL requests ?

1. select * from forge_upgrade_bucket WHERE script LIKE '/usr/share/codendi/plugins/agiledashboard/%'; (OR select * from forge_upgrade_bucket WHERE script LIKE '/usr/share/tuleap/plugins/agiledashboard/%'; if you are in centos 6)
2. select * from plugin_agiledashboard_configuration;

Hi Yannis,
yes, I run it and it returned with "System up-to-date". But I have the feeling that forgeupgrade does not work correctly on this machine because I did not update it for some weeks and I expected at least some updates. I have a second machine with Tuleap and on this machine forgeupgrade regularly modifies some tables.

I have executed forgeupgrade with option "already-applied" and got a lot of entries with "Skpped" (especially for agile dashboard). I suppose that this is not correct and probably the reason for the problem.