Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
10.5.13
-
None
Description
upgrade from Mariadb 10.5.10 to 10.5.13 results in [ERROR] InnoDB: corrupted TRX_NO
innodb_change_buffering=none
does not help to resolve corruption in undo. It does not happen all the times, but sporadically. But it's reproducible on the customer sites.
Not sure whether directly related but there is a reference to same error due os bug https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=257728
Attachments
Issue Links
- is caused by
-
MDEV-15132 Avoid accessing the TRX_SYS page
-
- Closed
-
- relates to
-
MDEV-15912 InnoDB: Failing assertion: purge_sys.tail.commit <= purge_sys.rseg->last_commit upon upgrade from 10.0 or 10.1 to 10.3
-
- Closed
-
-
MDEV-29111 locking occurring when upgrading to 10.5 from 10.1
-
- Closed
-
MDEV-26537was a regression that affected the 10.5.12 release when running on particular operating systems or file systems. I do not think that it should be relevant here.Does the database always restart correctly when using 10.5.10?