Resolved Crash after MariaDB 10.5.17 Update

[ggos]

Server operating system version

Centos 7

Plesk version and microupdate number

Plesk Obsidian 18.0.46

I have tried at my PLESK (Obsidian 18.0.46) to update my MariaDB-server version from 10.5.13 to 10.5.17 (via Plesk "System Updates" package upgrade) and during updating the connection to DB is lost and an PLESK Error page occurs with "Server Error: 500 Plesk\Exception\Database" with the message: "DB query failed: SQLSTATE[HY000] [2002] No such file or directory" at file "Mysql.php" on line 79.

After restoring a server backup I tried to upgrade mariadb manually by shell. The upgrade was successful, but afterwards I was not able to start mariadb with "systemctl start mariadb". This caused the error: "Job for mariadb.service failed because the control process exited with error code. See "systemctl status mariadb.service" and "journalctl -xe" for details.". The command "systemctl status mariadb" shows under Process: "(code=exited, status=1/FAILURE)".

What is going wrong with this MariaDB update? How can I solve this issue?

 

[Bitpalast]

It is correct that during the upgrade process the database connection breaks. This is by design, because the existing server and client are removed (this is when the connection breaks) and a new server and client are installed. The issue can normally be ignored, because it is automatically fixed once the new server is started.

The error message you quote is a general message that prompts you to investigate the root cause of the failure. You will normally find a detailed track record of the root cause in /var/log/messages (syslog etc. whatever your OS is). MariaDB logs all important steps when starting up and gives detailed hints why it cannot start. If you could paste these log parts here, likely other experts or me can help you sorting it out.

 

[Cike76]

ggos

Run systemctl status mariadb.service and paste the output here so we can see the error you are having...

 

[ggos]

Thank you very much for your help. I had a closer look into the Error Log file in /var/log/mariadb directory and it showed the error:

2022-09-06 9:16:36 0 [ERROR] InnoDB: innodb_page_size=16384 requires innodb_buffer_pool_size >= 5MiB current 2MiB

So I increased the innodb_buffer_pool_size in the /etc/my.cnf file and started the mariadb service again (systemctl start mariadb.service).
Now it's working fine.

Unfortunately, it is a bit tedious that this is not done automatically by PLESK...