Issue Since Update #1 on August 12, the databases are no longer accessible from outside. No other changes were made to the server. Restart and repair did no

[whu]

Server operating system version

Ubuntu 20.04.6 LTS

Plesk version and microupdate number

Plesk Obsidian Version 18.0.63 Update #2

Plesk Obsidian
Version 18.0.63 Update #2

Since Update #1 on August 12, the databases are no longer accessible from outside.
No other changes were made to the server. Restart and repair did not help.
The firewall could not be enabled properly but worked again after a restart.

I would appreciate any tips. Thank you!

 

[Sebahat.hadzhi]

Hello, @whu. We are sorry to hear that you are experiencing issues connecting to your databases. Could you please provide us with additional details, such as the database service you are using and any particular error you get when attempting to establish connection? If you are using MySQL/MariaDB that article might be useful.

 

[whu]

Hello, @whu. We are sorry to hear that you are experiencing issues connecting to your databases. Could you please provide us with additional details, such as the database service you are using and any particular error you get when attempting to establish connection? If you are using MySQL/MariaDB that article might be useful.

Thanks for your reply, but i get an 10061 Error. Think thats not like in the article!
yes it is a maria d

 

[whu]

We tried to set the STRICT mode in MariaDB to OFF!
As soon as we apply this, MariaDB crashes!
It can't be restarted, neither with OFF nor with ON.
Only a server restart helps.

Error: "Failed to start MariaDB 10.11.9 database server."

According to the update list, we should be running MariaDB 10.11.8!?
However, our Plesk shows that we have version 10.11.9 installed.
Only Plesk updates were applied.

The question now arises: Why do we encounter the error "Failed to start MariaDB 10.11.9 database server" on the server?

Plesk Obsidian 18.0.636 August 2024We are glad to present the “InnoDB repair” tool you can use to check for and attempt to automatically fix InnoDB corruption in MariaDB databases. InnoDB corruption is a serious issue, preventing the database server from working properly. Fixing it by hand requires time and expertise.You can access the tool by browsing https://:8443/repair/ and authenticating with the Plesk administrator credentials.
Updated MariaDB 10.11 to version 10.11.8.
Updated MariaDB 10.6 to version 10.6.18.
Updated MariaDB 10.5 to version 10.5.25.

 

[whu]

Mariadb start on Boot works but systemctl restart mariadb fails with Process: 7926 ExecStartPost=/etc/mysql/debian-start (code=exited, status=226/NAMESPACE)

 

[Bitpalast]

Let's have a look at what's been logged during the start attempt of the MariaDB service, e.g.
# journalctl -u mariadb.service --since today

 

[whu]

Aug 22 14:28:01 test.at mariadbd[689]: 2024-08-22 14:28:01 202014 [Warning] Aborted connection 202014 to db: 'psa' user: 'admin' host: 'localh>
Aug 22 18:42:54 test.at systemd[1]: Stopping MariaDB 10.11.9 database server...
Aug 22 18:42:54 test.at mariadbd[689]: 2024-08-22 18:42:54 0 [Note] /usr/sbin/mariadbd (initiated by: unknown): Normal shutdown
Aug 22 18:42:55 test.at mariadbd[689]: 2024-08-22 18:42:55 0 [Note] InnoDB: FTS optimize thread exiting.
Aug 22 18:42:55 test.at mariadbd[689]: 2024-08-22 18:42:55 0 [Note] InnoDB: Starting shutdown...
Aug 22 18:42:55 test.at mariadbd[689]: 2024-08-22 18:42:55 0 [Note] InnoDB: Dumping buffer pool(s) to /var/lib/mysql/ib_buffer_pool
Aug 22 18:42:55 test.at mariadbd[689]: 2024-08-22 18:42:55 0 [Note] InnoDB: Restricted to 2016 pages due to innodb_buf_pool_dump_pct=25
Aug 22 18:42:55 test.at mariadbd[689]: 2024-08-22 18:42:55 0 [Note] InnoDB: Buffer pool(s) dump completed at 240822 18:42:55
Aug 22 18:42:55 test.at mariadbd[689]: 2024-08-22 18:42:55 0 [Note] InnoDB: Removed temporary tablespace data file: "./ibtmp1"
Aug 22 18:42:55 test.at mariadbd[689]: 2024-08-22 18:42:55 0 [Note] InnoDB: Shutdown completed; log sequence number 1499258771203; transaction>
Aug 22 18:42:55 test.at mariadbd[689]: 2024-08-22 18:42:55 0 [Note] /usr/sbin/mariadbd: Shutdown complete
Aug 22 18:42:56 test.at systemd[1]: mariadb.service: Succeeded.
Aug 22 18:42:56 test.at systemd[1]: Stopped MariaDB 10.11.9 database server.
Aug 22 18:42:56 test.at systemd[1]: Starting MariaDB 10.11.9 database server...
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [ERROR] feedback plugin: failed to retrieve the MAC address
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] Starting MariaDB 10.11.9-MariaDB-ubu2004 source revision 0e8fb977b00983>
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Compressed tables use zlib 1.2.11
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Using transactional memory
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Number of transaction pools: 1
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Using crc32 + pclmulqdq instructions
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Using Linux native AIO
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Initializing buffer pool, total size = 128.000MiB, chunk size =>
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Completed initialization of buffer pool
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Buffered log writes (block size=512 bytes)
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: End of log at LSN=1499258771203
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: 128 rollback segments are active.
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Setting file './ibtmp1' size to 12.000MiB. Physically writing t>
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: File './ibtmp1' size is now 12.000MiB.
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: log sequence number 1499258771203; transaction id 1049642948
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Loading buffer pool(s) from /var/lib/mysql/ib_buffer_pool
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] Plugin 'FEEDBACK' is disabled.
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Warning] You need to use --log-bin to make --expire-logs-days or --binlog-exp>
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] Server socket created on IP: '0.0.0.0'.
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] Server socket created on IP: '::'.
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] /usr/sbin/mariadbd: ready for connections.
Aug 22 18:42:56 test.at mariadbd[692035]: Version: '10.11.9-MariaDB-ubu2004' socket: '/run/mysqld/mysqld.sock' port: 3306 mariadb.org binar>
Aug 22 18:42:56 test.at systemd[692050]: mariadb.service: Failed to set up mount namespacing: Cannot allocate memory
Aug 22 18:42:56 test.at systemd[692050]: mariadb.service: Failed at step NAMESPACE spawning /etc/mysql/debian-start: Cannot allocate memory
Aug 22 18:42:56 test.at systemd[1]: mariadb.service: Control process exited, code=exited, status=226/NAMESPACE
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] /usr/sbin/mariadbd (initiated by: unknown): Normal shutdown
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Buffer pool(s) load aborted due to user instigated abort at 240>
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Dumping of buffer pool not started as load was incomplete
Aug 22 18:42:56 test.at mariadbd[692035]: 2024-08-22 18:42:56 0 [Note] InnoDB: Starting shutdown...

 

[Bitpalast]

The relevant lines are:

Aug 22 18:42:56 test.at systemd[692050]: mariadb.service: Failed to set up mount namespacing: Cannot allocate memory
Aug 22 18:42:56 test.at systemd[692050]: mariadb.service: Failed at step NAMESPACE spawning /etc/mysql/debian-start: Cannot allocate memory
Aug 22 18:42:56 test.at systemd[1]: mariadb.service: Control process exited, code=exited, status=226/NAMESPACE

... but I have not encountered the exact issue before. I believe that this is not about RAM, but about a disk access issue or disk space issue. Potential further steps: Verify that SELinux allows access to the MariaDB files. If it set to "enforcing" test to change it to "permissive", reboot, try again to start the database server. If you are using Apparmor, temporarily disable it, then check service start again. Is the /var/lib/mysql directory a physical directory? Is a "datadir" defined in the MariaDB configuration and does it point to the correct /var/lib/mysql path?

I think it's a complicated problem, and I'd recommend to delegate it to official Plesk support. It may need a true MariaDB expert to solve and some investigation on your server.