MariaDB version 10.5.4 changes
I think we have a problem with the startup script for MariaDB.
https://mariadb.com/kb/en/mysqld_safe/ https://mariadb.com/kb/en/mysqlserver/ If I start with the rc.mysqld script I can no longer stop the service. Code:
/etc/rc.d/rc.mysqld start Quote:
Code:
/etc/rc.d/rc.mysqld stop Quote:
Code:
/usr/share/mysql/mysql.server start Quote:
Code:
/usr/share/mysql/mysql.server stop Quote:
And that's not all! From the log: Quote:
|
Tried it and see the same thing that you report.
|
I can confirm problem, but solution is easy.
Mariadb daemon has no longer name "mysqld" only "mariadbd". Change ~61 line in /etc/rc.d/rc.mysql from "killall mysqld" to "killall mariadbd", or apply this patch: Code:
--- rc.mysqld<->2020-06-25 09:44:36.063446186 +0200 |
Quote:
Code:
/etc/rc.d/rc.mysqld --help Quote:
Code:
/usr/share/mysql/mysql.server --help Quote:
There is already a patch proposal to avoid terminating other mysql instances at: https://www.linuxquestions.org/quest...ml#post6137742 I also recommend the following when migrating to a major version for MariaDB: 1. backup to databases; 2. uninstall previous version; 3. cleaning the remaining folders/files; 4. install the new version; 5. configuration mysql_install_db --user = mysql /usr/bin/mysql_secure_installation 6. test mysql -u root -p |
And that's not all!
From the log: Quote:
https://github.com/MariaDB/server/pull/1611 |
See https://mariadb.com/kb/en/using-the-s3-storage-engine/
Its probably good that it not load... To get rid of it hash in "plugin-load-add=ha_s3" line /etc/my.cnf.d/s3.cnf |
So MariaDB 10.5.4, comment on that line in /etc/my.cnf.d/s3.cnf and change the daemon name from "mysqld" to "mariadbd" in rc.mysqld?
Or maybe we stick to mariadb-10.4.13-x86_64-3 for now? |
Quote:
Change daemon name is Slackware issue and error with not loading "alpha quality" plugin is IMHO harmless. |
Quote:
I installed the new version on the test server but not yet on the production one. |
All times are GMT -5. The time now is 11:46 AM. |