Symptoms

When trying to restart MySQL, this error will appear in system logs (journalctl OR /var/log/messages):

Oct 07 09:38:18 server.test-001.net systemd[13595]: Failed at step NAMESPACE spawning /bin/sh: No space left on device

Oct 07 09:38:18 server.test-001.net systemd[1]: mariadb.service: control process exited, code=exited status=226

Oct 07 09:38:18 server.test-001.net systemd[1]: Failed to start MariaDB 10.3.24 database server.

Oct 07 09:38:18 server.test-001.net systemd[1]: Unit mariadb.service entered failed state.

Oct 07 09:38:18 server.test-001.net systemd[1]: mariadb.service failed.

Oct 07 09:38:18 server.test-001.net systemd[1]: Starting MariaDB 10.3.24 database server...

Oct 07 09:38:18 server.test-001.net systemd[13605]: Failed at step NAMESPACE spawning /bin/sh: No space left on device

Oct 07 09:38:18 server.test-001.net systemd[1]: mariadb.service: control process exited, code=exited status=226

Oct 07 09:38:18 server.test-001.net systemd[1]: Failed to start MariaDB 10.3.24 database server

Description

This usually means that the system or at least one of the storage devices on the system is running out of space. Here "space" could mean both normal block storage or inode usage. In this specific case when checking the storage situation on the server, I can see that the /tmp directory is at 100% inode usage:

 df -hi

Filesystem                  Inodes IUsed IFree IUse% Mounted on

devtmpfs                      3.7M  413  3.7M    1% /dev
tmpfs                          3.7M    2  3.7M    1% /dev/shm
tmpfs                          3.7M  1.4K  3.7M    1% /run
tmpfs                          3.7M    16  3.7M    1% /sys/fs/cgroup
/dev/mapper/centos-root        13M  254K  13M    2% /
/dev/sda1                      512K  353  512K    1% /boot
/dev/mapper/w32fast-homefast  100M  78K  100M    1% /homefast
/dev/mapper/w32slow-homeslow  250M  1.8M  249M    1% /homeslow
/dev/mapper/w32bak-bak        150M  605  150M    1% /backup
/dev/loop0                      66K   66K     0  100% /tmp
tmpfs                          3.7M    1  3.7M    1% /run/user/0

Please refer to this link for more information on the difference between normal storage and inode usage: 

Disk Usage Vs. Inodes Usage

Workaround

You need to clear some of the files on the problematic directory/file system/device and then attempt to restart MySQL again using this command:

/scripts/restartsrv_mysql --restart