MariaDB MariaDB 10.4.12, 10.3.22 and 10.2.31, 10.1.44 and 5.5.67 now available

Yesterday i updated to 10.4.12 from 10.4.11, and today i got the same issue twice.

MariaDB crashes, it starts after the crash, and then it gets stuck here:

Code:

[Note] InnoDB: Read redo log up to LSN=1410431322624

My only option was to force kill MariaDB process's, delete the two log files and start MariaDB again. It worked, and site was up without any bug.

After some hours it happened again. Now im trying to figure it out, if it was coincidence or not.

Journal is also being spammed with this messages:

Code:

Page [page id: space=6277, page number=30] log sequence number 1409649641062 is in the future! Current system log sequence number 1408897359163.
2020-02-03 23:08:48 37 [ERROR] InnoDB: Your database may be corrupt or you may have copied the InnoDB tablespace but not the InnoDB log files. Please refer to https://mariadb.com/kb/en/l

I checked for this error and its caused by the crash, and the only way to fix it, is to backup and then restore the database.

 

And today it happened the third time.

Its safe to revert to MariaDB 10.3? I know there was some big changes in 10.4.

Or i could also revert to 10.4.11, if that's possible.

 

Yes, i was thinking doing this:

1. Mysqldump the whole database
2. Remove MariaDB 10.4
3. Install MariaDB 10.3
4. Restore Mysql backup

I also remember some years ago here in centminmod, that we found a bug and we reverted to the previous minor version. But i don't remember how to do that.

 

Ok, thank you. I will try it first on a test server.

 

Just to let everyone know, if someone also got this issue.

Doing a mysqldump of the Databases(not including system databases) and restoring the backup, completely solved the issue. It was not necessary to revert to an old MariaDB version.