Mariadb103 available

yes when I released, I cleaned the metacache of the repository to save time, createrepo does the metadata each 30minutes on my server. maybe yum did not clean fast enough its cache

happy to see a good end :slight_smile:

Might also be the scl-utils - I will check


1 Like

With the original VM that the installation was not working

379  2020-04-15 15:26:39 yum remove *\rh-mariadb103\* -y
380  2020-04-15 15:28:18 rm -rf /var/opt/rh/rh-mariadb103/
381  2020-04-15 15:28:18 config delete rh-mariadb103-mariadb
382  2020-04-15 15:31:57 yum clean all --enablerepo=*

383  2020-04-15 15:32:21 date
384  2020-04-15 15:32:29 yum install nethserver-rh-mariadb103 --enablerepo=stephdl -y
385  2020-04-15 15:43:48 date
386  2020-04-15 15:44:06 mysql103

                                           NOT WORKING - took more then 10 minutes for the installation

387  2020-04-15 15:44:25 yum remove scl-utils

388  2020-04-15 15:46:30 yum remove *\rh-mariadb103\* -y
389  2020-04-15 15:46:47 rm -rf /var/opt/rh/rh-mariadb103/
390  2020-04-15 15:46:54 config delete rh-mariadb103-mariadb
391  2020-04-15 15:47:02 yum clean all --enablerepo=*

392  2020-04-15 15:47:29 yum install

393  2020-04-15 15:48:14 date
394  2020-04-15 15:48:17 yum install nethserver-rh-mariadb103 --enablerepo=stephdl -y
395  2020-04-15 15:49:42 mysql103

                                          WORKING - about 2 minutes for the installation

The SCL Utils is the same version as before:

# rpm -qa | grep scl |  grep util

What do you think ?

The removal/reinstallation of the SCL Utils ?

After cleanup - installation (update) of nethserver-stephdl before installation ?
But the cleanup should have clear the repo and the installation of MariaDB 10.3 should have updated it ?

Very “bizarre” isn’t it ?


Actually working on the ldap authentification of dolibarr, this is really bizarre, update field in the mysql db, then trigger a script each time that you create a user… this is bizarre :expressionless:

For MariaDB 10.3, is it possible that the "issue’ is related to the RPM key ?

The only difference I see beetwen the installation which is working and the one which is not, is getting the key.


1 Like

A post was split to a new topic: Dolibarr and LDAP authentication

Hi @stephdl,

after installing nethserver-mariadb103 for installing Drupal we get the same errors as @michelandre reported here including the long install time and service down. The error occurs on a fresh VM too.

The mysqld status shows the rh-mariadb103-mariadb status instead of the default mariadb 5?

[root@testserver ~]# systemctl status mysqld
● rh-mariadb103-mariadb.service - MariaDB 10.3 database server
   Loaded: loaded (/usr/lib/systemd/system/rh-mariadb103-mariadb.service; enabled; vendor preset: disabled)
   Active: failed (Result: exit-code) since Fri 2020-08-28 18:43:30 CEST; 2s ago
     Docs: man:mysqld(8)
  Process: 8127 ExecStartPre=/usr/bin/scl enable $RH_MARIADB103_SCLS_ENABLED -- /opt/rh/rh-mariadb103/root/usr/libexec/mysql-prepare-db-dir %n (code=exited, status=1/FAILURE)
  Process: 8099 ExecStartPre=/usr/bin/scl enable $RH_MARIADB103_SCLS_ENABLED -- /opt/rh/rh-mariadb103/root/usr/libexec/mysql-check-socket (code=exited, status=0/SUCCESS)
  Process: 8093 ExecStartPre=/usr/bin/scl enable $RH_MARIADB103_SCLS_ENABLED -- /usr/bin/scl_enabled rh-mariadb103 (code=exited, status=0/SUCCESS)
 Main PID: 2559
   CGroup: /system.slice/mysqld.service
           ├─2559 /bin/sh /usr/bin/mysqld_safe --basedir=/usr
           └─2734 /usr/libexec/mysqld --basedir=/usr --datadir=/var/lib/mysql --plugin-dir=/usr/lib64/mysql/plugin --log-error=/var/log/mariadb/mariadb.log --pid-file=/var/run/mariadb/ --socket=/var/lib/mysql/mysql.sock

Aug 28 18:43:29 testserver.domain.local systemd[1]: Starting MariaDB 10.3 database server...
Aug 28 18:43:30 testserver.domain.local systemd[1]: rh-mariadb103-mariadb.service: control process exited, code=exited status=1
Aug 28 18:43:30 testserver.domain.local systemd[1]: Failed to start MariaDB 10.3 database server.
Aug 28 18:43:30 testserver.domain.local systemd[1]: Unit rh-mariadb103-mariadb.service entered failed state.
Aug 28 18:43:30 testserver.domain.local systemd[1]: rh-mariadb103-mariadb.service failed.


Database MariaDB is not initialized, but the directory /var/opt/rh/rh-mariadb103/lib/mysql is not empty, so initialization cannot be done
Make sure the /var/opt/rh/rh-mariadb103/lib/mysql is empty before running mysql-prepare-db-dir

After deleting the directory and restarting it works:

rm /var/opt/rh/rh-mariadb103/lib/mysql/*
signal-event nethserver-rh-mariadb103-save

I compared installs of nethserver-rh-mariadb102 and 103 and I think the services are linked wrong, mysql and mysqld are somehow linked to rh-mariadb103-mariadb.

Nethserver-rh-mariadb103 creates 3 links on installation

Created symlink from /etc/systemd/system/mysql.service to /usr/lib/systemd/system/rh-mariadb103-mariadb.service.
Created symlink from /etc/systemd/system/mysqld.service to /usr/lib/systemd/system/rh-mariadb103-mariadb.service.
Created symlink from /etc/systemd/system/ to /usr/lib/systemd/system/rh-mariadb103-mariadb.service.

whereas Nethserver-rh-mariadb102 only creates 1:

Created symlink from /etc/systemd/system/ to /usr/lib/systemd/system/rh-mariadb102-mariadb.service.

Here are the wrong symlinks:

[root@testserver ~]# ls -lisa /etc/systemd/system/mysql*.service
33657745 0 lrwxrwxrwx 1 root root 53 Aug 28 18:22 /etc/systemd/system/mysqld.service -> /usr/lib/systemd/system/rh-mariadb103-mariadb.service
33657744 0 lrwxrwxrwx 1 root root 53 Aug 28 18:22 /etc/systemd/system/mysql.service -> /usr/lib/systemd/system/rh-mariadb103-mariadb.service

I don’t know where they come from…do you have an idea? I think the wrong symlinks are causing the bad initialization of the mariadb103.


tk will check ASAP

1 Like

first test for me it worked because mysql was not installed on my system but indeed I do not understand

Created symlink from /etc/systemd/system/mysql.service to /usr/lib/systemd/system/rh-mariadb103-mariadb.service.
Created symlink from /etc/systemd/system/mysqld.service to /usr/lib/systemd/system/rh-mariadb103-mariadb.service.

this is mine

Created symlink from /etc/systemd/system/ to /usr/lib/systemd/system/rh-mariadb103-mariadb.service.

need to check

1 Like

it comes directly from the systemd service /usr/lib/systemd/system/rh-mariadb103-mariadb.service


for /usr/lib/systemd/system/rh-mariadb102-mariadb.service we just have


I need to think the best way to solve it

1 Like

Fun I discovered a nice tool with systemd


to supervise which services has been customized

1 Like

would you mind to try it again with all version installed of mariadb (base, 101, 102, 103)

yum install nethserver-phpmyadmin


yum install nethserver-rh-mariadb10{1,2}

then install mariadb103

yum install

expand phpmyadmin to check if you can connect

signal-event nethserver-phpmyadmin-save

admin password is

[root@ns7dev9 ~]# cat .my.cnf

check the service states

 systemctl status rh-mariadb10*-mariadb.service
 systemctl status mysqld.service

It’s working now! :+1: Thanks for the quick fix!

Services are running without errors after install and I could login via phpmyadmin to every instance.

1 Like

Hi @stephdl,

To install Mariadb103, should install Mariadb101 first, then 102, or should install directly 103?
I have already installed Phpmyadmin from the your repo and Mariadb which came with NS.


I think you can only install mariadb103 and mysql base, I wanted to be sure that all the stack can work together

or only mariadb103 if needed

So, only the above command and should be ok, yes?

to install and enable rh-mariadb103, yes

1 Like

automatically done, released now


After this update “Updated: nethserver-rh-mariadb103-0.0.3-1.ns7.sdl.noarch”, everything works fine!

Thank you!

1 Like

Hello Stephdl,
We really appreciate your work.
Are you planning an upgrade from Mariadb 103 to Mariadb 105 soon ?
Is there a way to make it ourselves ?

MariaDB 10.5.8 Stable