Resolved Mariadb stop working in upgrade from Ubuntu 20.04 to 22.04

[rodrom]

Server operating system version

Ubuntu 22.04 Jammy

Plesk version and microupdate number

Plesk Obsidian 18.0.68.2

Hi,

I'm trying to upgrade ubuntu fosa (20.04) to jammy (22.04). But something went wrong in the middle of the process. I follow all the steps in a testing environment and it worked without problems. But, when I tried on the VPS, following the instructions for the manual step by step process from this website:

How to perform dist-upgrade procedure on Linux server with Plesk?

Applicable to: Plesk for Linux How to perform a dist-upgrade? Warning: The dist-upgrade procedure is complex, hence you may want to consider migrating instead as explained in Plesk Migration and...

support.plesk.com

The issue arose. After the step 18: reboot. mariadb doesnt start anymore. Executing step 21 (`plesk installer update --skip-cleanup`) doesnt work because mariadb is not able to start.

I did a full backup server before starting the process, but unfurtonately I was not able to do a snapshot of the VPS, because my provider doesnt allow it.

I dont know how to restore the backup, because as far as I understand, before that, I need mariadb to work.

Any suggestion?

Enclosed you can find several logs and outputs that can give you some better knowledge of the problem. As spoiler i give the output of systemctl status mariadb

Expand/Collapse Block: systemctl status mariadb

mariadb.service - MariaDB 10.6.21 database server
Loaded: loaded (/lib/systemd/system/mariadb.service; enabled; vendor preset: enabled)
Active: failed (Result: exit-code) since Wed 2025-04-16 19:30:07 UTC; 1s ago
Docs: man:mariadbd(8)
https://mariadb.com/kb/en/library/systemd/
Process: 56820 ExecStartPre=/usr/bin/install -m 755 -o mysql -g root -d /var/run/mysqld (code=exited, status=0/SUCCESS)
Process: 56821 ExecStartPre=/bin/sh -c systemctl unset-environment _WSREP_START_POSITION (code=exited, status=0/SUCCESS)
Process: 56823 ExecStartPre=/bin/sh -c [ ! -e /usr/bin/galera_recovery ] && VAR= || VAR=`/usr/bin/galera_recovery`; [ $? -eq 0 ] && systemctl set-environment _WSREP_START_POSITION=$VAR || exit 1 (code=exited, status=0/SUCCESS)
Process: 56861 ExecStart=/usr/sbin/mariadbd $MYSQLD_OPTS $_WSREP_NEW_CLUSTER $_WSREP_START_POSITION (code=exited, status=218/CAPABILITIES)
Main PID: 56861 (code=exited, status=218/CAPABILITIES)

Apr 16 19:30:07 example.com systemd[1]: Starting MariaDB 10.6.21 database server...
Apr 16 19:30:07 example.com systemd[56861]: mariadb.service: Failed to apply ambient capabilities (before UID change): Operation not permitted
Apr 16 19:30:07 example.com systemd[56861]: mariadb.service: Failed at step CAPABILITIES spawning /usr/sbin/mariadbd: Operation not permitted
Apr 16 19:30:07 example.com systemd[1]: mariadb.service: Main process exited, code=exited, status=218/CAPABILITIES
Apr 16 19:30:07 example.com systemd[1]: mariadb.service: Failed with result 'exit-code'.
Apr 16 19:30:07 example.com systemd[1]: Failed to start MariaDB 10.6.21 database server.

 

[rodrom]

I was able to fix the problem asking chatgpt... that shows two things: that Bing is totally broken, and my knowledge of Mariadb and Systemd is not the best.

Anyway, this is the link to the conversation:

ChatGPT - VPS Plesk Troubleshooting Help

Shared via ChatGPT

chatgpt.com

Steps to solve this problem.

1. Override the configuration of mariadb

sudo systemctl edit mariadb

[Service]
CapabilityBoundingSet=
AmbientCapabilities=

2. Reload systemd and restart Mariadb

sudo systemctl daemon-reexec
sudo systemctl daemon-reload
sudo systemctl restart mariadb

3. Check status

sudo systemctl status mariadb

Continue with the step 21.

I was able to finish the upgrade without issues.