MySQL InnoDB 问题 - InnoDB 注册失败

Question

I recently attempted to upgrade a MySQL 5.1 server to 5.7. When the server wouldn't start, I discovered you have to export the data first, else do numerous upgrades (the binaries for which aren't available anymore), so I rolled back to 5.1 to do an export.

The problem is, back on 5.1, InnoDB will no longer register. In the error log i get:

InnoDB: Error: log file ./ib_logfile0 is of different size 0 50331648 bytes
InnoDB: than specified in the .cnf file 0 5242880 bytes!
160822 17:05:12 [ERROR] Plugin 'InnoDB' init function returned error.
160822 17:05:12 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed.

As a workaround I set innodb_log_file_size=50331648 in my.cnf . Another restart and:

InnoDB: No valid checkpoint found.
InnoDB: If this error appears when you are creating an InnoDB database,
InnoDB: the problem may be that during an earlier attempt you managed
InnoDB: to create the InnoDB data files, but log file creation failed.
InnoDB: If that is the case, please refer to
InnoDB: http://dev.mysql.com/doc/refman/5.1/en/error-creating-innodb.html

Reading that links it suggests to delete all files created by InnoDB: all ibdata files and all ib_logfile files . I have a 120 GB ibdata file that I most certainly do not intend to delete, so this is a non-starter. I did try renaming the logfiles though and the server still wouldn't come up.

Any pointers from here?

Edit: I also tried renaming the iblog files. That causes another error where mysql suggests setting innodb_force_recovery=6 . Doing that causes these errors:

160823 12:17:32  InnoDB: Page checksum 271832187, prior-to-4.0.14-form checksum 315921779
InnoDB: stored checksum 401867329, prior-to-4.0.14-form stored checksum 401867329
InnoDB: Page lsn 96 3891045697, low 4 bytes of lsn at page end 3891045697
InnoDB: Page number (if stored to page already) 966706,
InnoDB: space id (if created with >= MySQL-4.1.1 and stored already) 0
InnoDB: Page may be a system page
InnoDB: Database page corruption on disk or a failed
InnoDB: file read of page 966706.
InnoDB: You may have to recover from a backup.
InnoDB: It is also possible that your operating
InnoDB: system has corrupted its own file cache
InnoDB: and rebooting your computer removes the
InnoDB: error.
InnoDB: If the corrupt page is an index page
InnoDB: you can also try to fix the corruption
InnoDB: by dumping, dropping, and reimporting
InnoDB: the corrupt table. You can use CHECK
InnoDB: TABLE to scan your table for corruption.
InnoDB: See also http://dev.mysql.com/doc/refman/5.1/en/forcing-innodb-recovery.html
InnoDB: about forcing recovery.
InnoDB: Database page corruption on disk or a failed
InnoDB: file read of page 966707.
InnoDB: You may have to recover from a backup.
160823 12:17:32  InnoDB: Page dump in ascii and hex (16384 bytes):

Answer 1

I encounter the similar problem just now. I run both MySQL 5.1 and 5.7 in the same server.

With MySQL 5.1, the default innodb_log_file_size is 5m.

mysql> show variables like 'innodb_log_file_size';

    +----------------------+---------+
    | Variable_name        | Value   |
    +----------------------+---------+
    | innodb_log_file_size | 5242880 |
    +----------------------+---------+
    1 row in set (0.00 sec)

While MySQL 5.7, the default innodb_log_file_size is 48m.

mysql> show variables like 'innodb_log_file_size';
+----------------------+----------+
| Variable_name        | Value    |
+----------------------+----------+
| innodb_log_file_size | 50331648 |
+----------------------+----------+
1 row in set (0.01 sec)

When upgrade a MySQL 5.1 server to 5.7 and then downgrade to 5.1, the ib_logfile0/1 file was changed in a magic way. The checkpoint of log file cannot be identified. So innodb cannot execute a normal recovery using the log files. I didn't make a deep research so i can't give a detailed explanation.

Solution

Backup datafile and restart MySQL

mv ibdata1 ibdata1.bak
mv ib_logfile0 ib_logfile0.bak
mv ib_logfile1 ib_logfile1.bak
service mysqld restart

Modify my.cnf file, add the following line

innodb_force_recovery=6

Restore datafile and restart again

mv ibdata1.bak ibdata1

service mysqld restart

Now you will see MySQL start successfully. Ibdata1 may not work properly. You can backup your application data using mysqldump. Then remove the recovery line in my.cnf file. Remove ibdata and logfile as well. Restart MySQL and Import your backup.

A little complex, but these help me.

Faults

As the redo log file was recreated, a little piece of log cannot be redo.

Answer 2

Do not delete ibdata* . That is where all the data is!

By now, the log files should not matter. So, there are a couple of ways to deal with the 'error'.

Get back the old my.cnf . If that is not practical, at least set

innodb_log_file_size = 50331648 

(The change to the log file is a bit tricky. Apparently 5.7 increased it to the new default of 48MB without a problem. However, 5.1 does not have the code to dynamically change the file size.)

Another approach is to delete (or move aside) the iblog* files.