The root cause is the WAL logging of file operation when the actual
operation fails afterwards. It creates a situation with a log entry for
a operation that would always fail. I could simulate both the backup
scenario error and Innodb recovery failure exploiting the weakness.
We are following WAL for file rename operation and once logged the
operation must eventually complete successfully, or it is a major
catastrophe. Right now, we fail for rename and handle it as normal error
and it is the problem.
I created a patch to address RENAME operation to a non existing schema
where the destination schema directory is missing. The patch checks for
the missing schema before logging in an attempt to avoid the failure
after WAL log is written/flushed. I also checked that the schema cannot
be dropped or there cannot be any race with other rename to the same
file. This is protected by the MDL lock in SQL today.
The patch should this be a good improvement over the current situation
and solves the issue at hand.
When attempting to rename a table to a non-existing database,
InnoDB would misleadingly report "OS error 71" when in fact the
error code is InnoDB's own (OS_FILE_NOT_FOUND), and not report
both pathnames. Errors on rename could occur due to reasons
connected to either pathname.
os_file_handle_rename_error(): New function, to report errors in
renaming files.
Cherry-pick the commit from MySQL 5.7.19, and adapt the test case:
commit 45c933ac19c73a3e9c756a87ee1ba18ba1ac564c
Author: Aakanksha Verma <aakanksha.verma@oracle.com>
Date: Tue Mar 21 10:31:43 2017 +0530
Bug #25189192 ERRORS WHEN RESTARTING MYSQL AFTER RENAME TABLE.
PROBLEM
While renaming table innodb doesn't update the InnoDB Dictionary table
INNODB_SYS_DATAFILES incase there is change in database while doing
rename table. Hence on a restart the server log shows error that it
couldnt find table with old path before rename which has actually been
renamed. So the errors would only vanish if we update the system
tablespace
FIX
Update the innodb dictionary table with new path in the case there is
not a change in the table but the database holding the table as well.
Reviewed-by: Jimmy Yang<Jimmy.Yang@oracle.com>
RB: 15751
