Just updated mariadb from 10.5.11 to 10.5.12 this broke akonadi and so no email

Did anybody else experienced the same issue: just updated mariadb from 10.5.11 to 10.5.12 this broke akonadi and so no email??

Downgrading back to mariadb-10.5.11 made it all work again.

It’s the mariadb from the repo-sle-update repository.

What’s the openSUSE Leap 15.3 location to report bugs?

I am runnig 15.2, thus I can not check easily, but Is that newer version coming from an Update repo?
When yes, it sounds as worth a bug report. Starnge however that others did not report it.

https://bugzilla.opensuse.org/index.cgi
Same username/password as here.

Thanks. I included the repository in the initial message: repo-sle-update repository.

Created: https://bugzilla.opensuse.org/show_bug.cgi?id=1190191

I am not sure you can add SLE repos without impunity.

Leap 15.3 is based on SLE (they have the same base).
As far as I know, these are the repositories for 15.3.

As said, I run 15.2. But I got the impression that the SLE repos needed for 15.3 are added during the uograde process. Again, I am not sure if adding more spontanious is supported (e.g. by the testing). But maybe a 15.3 specialist will tune in.

Thats one of the Repos added by upgrading from Leap 15.2 to Leap 15.3, and you get from this Repos important updates such as Kernel-updates or now the mariadb update…

l  | kernel-default-base         | Paket      | 5.3.18-59.19.1.18.10.1 | x86_64 | Update repository with updates from SUSE Linux Enterprise 15
i+ | mariadb                     | Paket      | 10.5.12-3.6.1          | x86_64 | Update repository with updates from SUSE Linux Enterprise 15

Then it should already have been on his installation? He says he “created” it.

In any case, when it is part of the standard repos for 15.3 (rather confusing so), then a bug report is OK.

Kmail and akonadi work without problems on my 15.3 machine after the mariadb 10.5.12-3.6.1 update last week.


i+ | akonadi-server        | package | 20.04.2-bp153.4.2.1 | x86_64 | Update repository of openSUSE Backports
i+ | mariadb               | package | 10.5.12-3.6.1       | x86_64 | Update repository with updates from SUSE Linux Enterprise 15 


Could you can check if the journal had some akonadi error messages after the update ?

Which journal are you referring to? I checked /var/log/messages (journald) and there is no message from Akonadi, that’shelpfull. It just states that it can’t connect, but that we already know.

The log shows that mariadb (…11) was succesfully replaces with (…12).

Perhaps this is something:

2021-09-04T11:15:54.139807+02:00  akonadiserver[2595]: org.kde.pim.akonadiserver: Starting up the **Akonadi** Server...
2021-09-04T11:15:58.262640+02:00  org.freedesktop.**Akonadi**.Control[2619]: libEGL warning: DRI2: failed to authenticate
2021-09-04T11:15:59.515880+02:00 akonadiserver[2595]: org.kde.pim.akonadiserver: Shutting down **Akonadi**Server...

The update is now on my second 15.3 machine and again no problem with mariadb 10.5.12 and Kmail/akonadi here.
You could try to start akonadi manual with the --verbose option


akonadictl --verbose start 

maybe this will show some helpful info. You could also check if there are errors in ~/.local/share/akonadi/db_data/mysql.err

Thanks for the good hints/tips!

This is from mysql.err


2021-09-04 11:12:50 0 [ERROR] Fatal error: Can't open and lock privilege tables: Table 'mysql.db' doesn't exist
2021-09-04 11:12:50 0 [ERROR] Can't open and lock privilege tables: Table 'mysql.db' doesn't exist
2021-09-04 11:12:50 0 [Note] /usr/sbin/mysqld (initiated by: unknown): Normal shutdown
2021-09-04 11:12:50 0 [Note] InnoDB: FTS optimize thread exiting.
2021-09-04 11:12:50 0 [Note] InnoDB: Starting shutdown...
2021-09-04 11:12:50 0 [Note] InnoDB: Dumping buffer pool(s) to /home/userjoe/.local/share/akonadi/db_data/ib_buffer_pool
2021-09-04 11:12:50 0 [Note] InnoDB: Restricted to 2016 pages due to innodb_buf_pool_dump_pct=25
2021-09-04 11:12:50 0 [Note] InnoDB: Buffer pool(s) dump completed at 210904 11:12:50
2021-09-04 11:12:51 0 [Note] InnoDB: Removed temporary tablespace data file: "ibtmp1"
2021-09-04 11:12:51 0 [Note] InnoDB: Shutdown completed; log sequence number 53112589359; transaction id 5495884975336692350
2021-09-04 11:12:51 0 [Note] /usr/sbin/mysqld: Shutdown complete

2021-09-04 11:15:45 0 [Note] InnoDB: Uses event mutexes
2021-09-04 11:15:45 0 [Note] InnoDB: Compressed tables use zlib 1.2.11
2021-09-04 11:15:45 0 [Note] InnoDB: Number of pools: 1
2021-09-04 11:15:45 0 [Note] InnoDB: Using SSE4.2 crc32 instructions
2021-09-04 11:15:45 0 [Note] mysqld: O_TMPFILE is not supported on /tmp (disabling future attempts)
2021-09-04 11:15:45 0 [Note] InnoDB: Using Linux native AIO
2021-09-04 11:15:45 0 [Note] InnoDB: Initializing buffer pool, total size = 134217728, chunk size = 134217728
2021-09-04 11:15:45 0 [Note] InnoDB: Completed initialization of buffer pool
2021-09-04 11:15:46 0 [ERROR] InnoDB: corrupted TRX_NO 4c454e44415999b9
2021-09-04 11:15:46 0 [Note] InnoDB: Retry with innodb_force_recovery=5
2021-09-04 11:15:46 0 [ERROR] InnoDB: Plugin initialization aborted with error Data structure corruption
2021-09-04 11:15:46 0 [Note] InnoDB: Starting shutdown...
2021-09-04 11:15:46 0 [ERROR] Plugin 'InnoDB' init function returned error.
2021-09-04 11:15:46 0 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed.
2021-09-04 11:15:46 0 [Note] Plugin 'FEEDBACK' is disabled.
2021-09-04 11:15:46 0 [ERROR] Unknown/unsupported storage engine: innodb
2021-09-04 11:15:46 0 [ERROR] Aborting

It looks like a problem with the innodb engine.

The log lines just before the above lines, or the 10.5.11 engine:

2021-09-04  8:41:41 0 [Note] InnoDB: Uses event mutexes
2021-09-04  8:41:41 0 [Note] InnoDB: Compressed tables use zlib 1.2.11
2021-09-04  8:41:41 0 [Note] InnoDB: Number of pools: 1
2021-09-04  8:41:41 0 [Note] InnoDB: Using SSE4.2 crc32 instructions
2021-09-04  8:41:41 0 [Note] mysqld: O_TMPFILE is not supported on /tmp (disabling future attempts)
2021-09-04  8:41:41 0 [Note] InnoDB: Using Linux native AIO
2021-09-04  8:41:41 0 [Note] InnoDB: Initializing buffer pool, total size = 134217728, chunk size = 134217728
2021-09-04  8:41:41 0 [Note] InnoDB: Completed initialization of buffer pool
2021-09-04  8:41:41 0 [Note] InnoDB: 128 rollback segments are active.
2021-09-04  8:41:41 0 [Note] InnoDB: Creating shared tablespace for temporary tables
2021-09-04  8:41:41 0 [Note] InnoDB: Setting file './ibtmp1' size to 12 MB. Physically writing the file full; Please wait ...
2021-09-04  8:41:41 0 [Note] InnoDB: File './ibtmp1' size is now 12 MB.
2021-09-04  8:41:41 0 [Note] InnoDB: 10.5.11 started; log sequence number 53111891541; transaction id 5495884975336691524
2021-09-04  8:41:41 0 [Note] Plugin 'FEEDBACK' is disabled.
2021-09-04  8:41:41 0 [Note] InnoDB: Loading buffer pool(s) from /home/userjoe/.local/share/akonadi/db_data/ib_buffer_pool
2021-09-04  8:41:41 0 [ERROR] Can't open and lock privilege tables: Table 'mysql.servers' doesn't exist
2021-09-04  8:41:41 0 [Warning] Can't open and lock time zone table: Table 'mysql.time_zone_leap_second' doesn't exist trying to live without them
2021-09-04  8:41:41 1 [Warning] Failed to load slave replication state from table mysql.gtid_slave_pos: 1017: Can't find file: './mysql/' (errno: 2 "No such file or 
directory")
2021-09-04  8:41:41 0 [Note] Reading of all Master_info entries succeeded
2021-09-04  8:41:41 0 [Note] Added new Master_info '' to hash table
2021-09-04  8:41:41 0 [Note] /usr/sbin/mysqld: ready for connections.
Version: '10.5.11-MariaDB'  socket: '/run/user/1000/akonadi/mysql.socket'  port: 0  MariaDB package
2021-09-04  8:41:42 4 [ERROR] InnoDB: Table `mysql`.`innodb_table_stats` not found.
2021-09-04  8:41:42 0 [Note] InnoDB: Buffer pool(s) load completed at 210904  8:41:42

There seems to be an issue, but it starts at least.

Any idea what’s going on?

I’d say this is the culprit somehow:

I can’t actually help nor trace this as I am on 15.2. I can’t find any hint to “mysql.db” in any way, no hint in any error nor log nor config file. Not even

zypper se --provides mysql.db

provides any result. I should mention akonadi / kontact works without problems. Not sure about the changes from 15.2 to 15.3 nor about mariadb from 10.5.11 to 10.5.12. It seems to call a database (mysql.db) which isn’t there (resp. where it’s supposed to be). So is this call wrong or should there be a “mysql.db” somewhere? Or is this even completely irrelevant?

Thanks for your thinking out load, appreciated and just what I need. I’ll look into during the weekend, when I’ve a bit more time. Perhaps in the meantime some else may have a suggestion???

Does anybody how to initialize a mysql/mariadb database to be used with/for akonadi?

I know that I can do this with:


mysql_install_db --defaults-file=~/.local/share/akonadi/mysql.conf --datadir=/var/tmp/mariadb/joe/db_data

This will give me a DB engine, with only the information_schema database.

In an original akonadi database, there are also the mysql and performance_schema databases.

In short I expect this:

MariaDB (none)]> show databases;
+--------------------+
| Database           |
+--------------------+
| akonadi            |
| information_schema |
| mysql              |
| performance_schema |
+--------------------+

But I get:

MariaDB (none)]> show databases;
+--------------------+
| Database           |
+--------------------+
| information_schema |
+--------------------+

akonadi is of course missing, that is a database to import. But how or when do the other 2 (mysql and performance_schema) appear?

Does anyone know?

As you do not show your complete command line, the prompts are missing. Thus we can not see who (main interest root or normal user) is running all these commands. I hope you are aware of the fact the the Akonadi database is per (KDE) user and that thus using or not of a MariaDB on the system level has no connection to them. This is as I understand this.

It is for a regular user.

Indeed akonadi is for the user using KDE, I know. The thing is, when I update mariadb to the next version it breaks my current setup and I’ve to restore the databases (for various) users. That’s why I need to re-install the databases, and I try to do this in the (KDE) correct way.

If someone knows the script or executable that creates the akonadi DB, that would be nice to know as well.

It could of course be that people here know how to help you. But maybe you should go to a more specialized place.

There is the KDE forums: https://forum.kde.org/

And there is the KDE-PIM mailing list: kdepim-users@kde.org