Description:
* Create databases(db1, db2), tables, and load some data
* Execute backup using innobackup
* Drop the databases db1, db2
* Execute innobackup with --apply-logs
* Now check the backup directory created by innobackup and make some changes in the mysql/host.frm file (that is manually corrupt the file)
* Shutdown mysql server and then do innobackup with --copy-back (and it passes)
* Now restart the server and we will notice that server shuts down because of host.frm file.
Ideally corrupted file can result in restore failure(that is innobakcup --copy-back operation can fail) but can't lead to shutdown of server. This is a negative test and the expected result should be "restore failure".
How to repeat:
mysql> CREATE DATABASE db1;
Query OK, 1 row affected (0.00 sec)
mysql> CREATE DATABASE db2;
Query OK, 1 row affected (0.00 sec)
mysql> CREATE TABLE db1.t1(id INT)ENGINE=INNODB;
CREATE TABLE db2.t2(a CHAR(10))ENGINE=INNODB;
Query OK, 0 rows affected (0.12 sec)
mysql> CREATE TABLE db2.t2(a CHAR(10))ENGINE=INNODB;
Query OK, 0 rows affected (0.11 sec)
mysql>
mysql> INSERT INTO db1.t1 VALUES(1),(2),(3),(4);
Query OK, 4 rows affected (0.00 sec)
Records: 4 Duplicates: 0 Warnings: 0
mysql> INSERT INTO db2.t2 VALUES('a'),('b'),('c'),('d');
Query OK, 4 rows affected (0.00 sec)
Records: 4 Duplicates: 0 Warnings: 0
mysql>
mysql> SELECT * FROM db1.t1;
SELECT * FROM db2.t2;
+------+
| id |
+------+
| 1 |
| 2 |
| 3 |
| 4 |
+------+
4 rows in set (0.00 sec)
mysql> SELECT * FROM db2.t2;
+------+
| a |
+------+
| a |
| b |
| c |
| d |
+------+
4 rows in set (0.00 sec)
[hs221732@fimafeng10]/export/home2/tmp/mysql-5.1-ihb:> perl innobackup hema-inno.cnf mysql-test/backup/ --socket=/export/home2/tmp/mysql-5.1-ihb/mysql-test/var/tmp/mysqld.1.sock --user=root --ibbackup=/export/home2/tmp/mysql-5.1-ihb/install/ibbackup
mysql> drop database db1;
Query OK, 1 row affected (0.04 sec)
mysql> drop database db2;
Query OK, 1 row affected (0.07 sec)
Now manually corrupt the host.frm file in 2010-04-28_22-27-02/mysql/ directory
perl innobackup --apply-log hema-inno.cnf /export/home2/tmp/mysql-5.1-ihb/mysql-test/backup/2010-04-28_22-27-02/ --ibbackup=/export/home2/tmp/mysql-5.1-ihb/ibbackup
shutdown the server
./innobackup --copy-back hema-inno.cnf /export/home2/tmp/mysql-5.1-ihb/mysql-test/backup/2010-04-28_22-27-02/
Restart the server
Server log is:
100428 23:31:06 [Warning] '--default-character-set' is deprecated and will be removed in a future release. Please use '--character-set-server' instead.
100428 23:31:06 [Note] Plugin 'FEDERATED' is disabled.
100428 23:31:06 [Note] Plugin 'ndbcluster' is disabled.
InnoDB: The log file was created by ibbackup --apply-log at
InnoDB: ibbackup 100428 22:28:08
InnoDB: NOTE: the following crash recovery is part of a normal restore.
InnoDB: The log sequence number in ibdata files does not match
InnoDB: the log sequence number in the ib_logfiles!
100428 23:31:06 InnoDB: Database was not shut down normally!
InnoDB: Starting crash recovery.
InnoDB: Reading tablespace information from the .ibd files...
InnoDB: Restoring possible half-written data pages from the doublewrite
InnoDB: buffer...
100428 23:31:06 InnoDB: Started; log sequence number 0 53772
100428 23:31:06 [ERROR] /export/home2/tmp/mysql-5.1-ihb/sql/mysqld: Incorrect information in file: './mysql/host.frm'
100428 23:31:06 [ERROR] Fatal error: Can't open and lock privilege tables: Incorrect information in file: './mysql/host.frm'
100428 23:31:06 [Note] Got signal 15 to shutdown mysqld
As we can notice that server shuts down and gets signal 15 which is really a bad scenario.
Suggested fix:
Ideally the server should not shutdown and instead result in "restore failure" when innobackup --apply-log or innobackup --copy-back operation is performed.
Description: * Create databases(db1, db2), tables, and load some data * Execute backup using innobackup * Drop the databases db1, db2 * Execute innobackup with --apply-logs * Now check the backup directory created by innobackup and make some changes in the mysql/host.frm file (that is manually corrupt the file) * Shutdown mysql server and then do innobackup with --copy-back (and it passes) * Now restart the server and we will notice that server shuts down because of host.frm file. Ideally corrupted file can result in restore failure(that is innobakcup --copy-back operation can fail) but can't lead to shutdown of server. This is a negative test and the expected result should be "restore failure". How to repeat: mysql> CREATE DATABASE db1; Query OK, 1 row affected (0.00 sec) mysql> CREATE DATABASE db2; Query OK, 1 row affected (0.00 sec) mysql> CREATE TABLE db1.t1(id INT)ENGINE=INNODB; CREATE TABLE db2.t2(a CHAR(10))ENGINE=INNODB; Query OK, 0 rows affected (0.12 sec) mysql> CREATE TABLE db2.t2(a CHAR(10))ENGINE=INNODB; Query OK, 0 rows affected (0.11 sec) mysql> mysql> INSERT INTO db1.t1 VALUES(1),(2),(3),(4); Query OK, 4 rows affected (0.00 sec) Records: 4 Duplicates: 0 Warnings: 0 mysql> INSERT INTO db2.t2 VALUES('a'),('b'),('c'),('d'); Query OK, 4 rows affected (0.00 sec) Records: 4 Duplicates: 0 Warnings: 0 mysql> mysql> SELECT * FROM db1.t1; SELECT * FROM db2.t2; +------+ | id | +------+ | 1 | | 2 | | 3 | | 4 | +------+ 4 rows in set (0.00 sec) mysql> SELECT * FROM db2.t2; +------+ | a | +------+ | a | | b | | c | | d | +------+ 4 rows in set (0.00 sec) [hs221732@fimafeng10]/export/home2/tmp/mysql-5.1-ihb:> perl innobackup hema-inno.cnf mysql-test/backup/ --socket=/export/home2/tmp/mysql-5.1-ihb/mysql-test/var/tmp/mysqld.1.sock --user=root --ibbackup=/export/home2/tmp/mysql-5.1-ihb/install/ibbackup mysql> drop database db1; Query OK, 1 row affected (0.04 sec) mysql> drop database db2; Query OK, 1 row affected (0.07 sec) Now manually corrupt the host.frm file in 2010-04-28_22-27-02/mysql/ directory perl innobackup --apply-log hema-inno.cnf /export/home2/tmp/mysql-5.1-ihb/mysql-test/backup/2010-04-28_22-27-02/ --ibbackup=/export/home2/tmp/mysql-5.1-ihb/ibbackup shutdown the server ./innobackup --copy-back hema-inno.cnf /export/home2/tmp/mysql-5.1-ihb/mysql-test/backup/2010-04-28_22-27-02/ Restart the server Server log is: 100428 23:31:06 [Warning] '--default-character-set' is deprecated and will be removed in a future release. Please use '--character-set-server' instead. 100428 23:31:06 [Note] Plugin 'FEDERATED' is disabled. 100428 23:31:06 [Note] Plugin 'ndbcluster' is disabled. InnoDB: The log file was created by ibbackup --apply-log at InnoDB: ibbackup 100428 22:28:08 InnoDB: NOTE: the following crash recovery is part of a normal restore. InnoDB: The log sequence number in ibdata files does not match InnoDB: the log sequence number in the ib_logfiles! 100428 23:31:06 InnoDB: Database was not shut down normally! InnoDB: Starting crash recovery. InnoDB: Reading tablespace information from the .ibd files... InnoDB: Restoring possible half-written data pages from the doublewrite InnoDB: buffer... 100428 23:31:06 InnoDB: Started; log sequence number 0 53772 100428 23:31:06 [ERROR] /export/home2/tmp/mysql-5.1-ihb/sql/mysqld: Incorrect information in file: './mysql/host.frm' 100428 23:31:06 [ERROR] Fatal error: Can't open and lock privilege tables: Incorrect information in file: './mysql/host.frm' 100428 23:31:06 [Note] Got signal 15 to shutdown mysqld As we can notice that server shuts down and gets signal 15 which is really a bad scenario. Suggested fix: Ideally the server should not shutdown and instead result in "restore failure" when innobackup --apply-log or innobackup --copy-back operation is performed.