一、使用binlog日志
目标：
利用binlog恢复库表，要求如下：
- 启用binlog日志
- 创建db1库tb1表，插入3条记录
- 删除tb1表中刚插入的3条记录
- 使用mysqlbinlog恢复删除的3条记录
步骤：
步骤一：启用binlog日志 //记录除了查询之外的SQL命令，默认未启用
1）调整/etc/my.cnf配置，并重启服务
[root@dbsvr1 ~]# vim /etc/my.cnf
[mysqld]
.. ..
log-bin-index=mysql-bin //启用二进制日志，并指定前缀
server_id=1
binlog_format=STATEMENT //在Mysql5.7中，binlog日志格式默认为ROW，但它不记录sql语句上下文相关信息。需要将binlog日志格式修改为STATEMENT
.. ..
[root@dbsvr1 ~]# systemctl restart mysqld.service
2）确认binlog日志文件
新启用binlog后，每次启动MySQl服务都会新生成一份日志文件：
[root@dbsvr1 ~]# ls /var/lib/mysql/mysql-bin.*
/var/lib/mysql/mysql-bin.000001 /var/lib/mysql/mysql-bin.index
其中mysql-bin.index文件记录了当前保持的二进制文件列表：
[root@dbsvr1 ~]# cat /var/lib/mysql/mysql-bin.index
./mysql-bin.000001
重启MySQL服务程序，或者执行SQL操作“FLUSH LOGS;”，会生成一份新的日志：
[root@dbsvr1 ~]# ls /var/lib/mysql/mysql-bin.*
/var/lib/mysql/mysql-bin.000001 /var/lib/mysql/mysql-bin.index
/var/lib/mysql/mysql-bin.000002
[root@dbsvr1 ~]# cat /var/lib/mysql/mysql-bin.index
./mysql-bin.000001
./mysql-bin.000002
步骤二：利用binlog日志重做数据库操作
1）执行数据库表添加操作
创建db1·库tb1表，表结构自定义：
mysql> CREATE DATABASE db1;
Query OK, 1 row affected (0.05 sec)
mysql> USE db1;
Database changed
mysql> CREATE TABLE tb1(
-> id int(4) NOT NULL,name varchar(24)
-> );
Query OK, 0 rows affected (0.28 sec)
插入3条表记录：
mysql> INSERT INTO tb1 VALUES
-> (1,'Jack'),
-> (2,'Kenthy'),
-> (3,'Bob');
Query OK, 3 rows affected (0.12 sec)
Records: 3 Duplicates: 0 Warnings: 0
确认插入的表记录数据：
mysql> SELECT * FROM tb1;
+----+--------+
| id | name |
+----+--------+
| 1 | Jack |
| 2 | Kenthy |
| 3 | Bob |
+----+--------+
3 rows in set (0.00 sec)
2）删除前一步添加的3条表记录
执行删除所有表记录操作：
mysql> DELETE FROM tb1;
Query OK, 3 rows affected (0.09 sec)
确认删除结果：
mysql> SELECT * FROM tb1;
Empty set (0.00 sec)
步骤三：通过binlog日志恢复表记录
binlog会记录所有的数据库、表更改操作，所以可在必要的时候重新执行以前做过的一部分数据操作，但对于启用binlog之前已经存在的库、表数据将不适用。
根据上述“恢复被删除的3条表记录”的需求，应通过mysqlbinlog工具查看相关日志文件，找到删除这些表记录的时间点，只要恢复此前的SQL操作（主要是插入那3条记录的操作）即可。
1）查看mysql-bin.000002日志内容
[root@dbsvr1 ~]# mysqlbinlog /var/lib/mysql/mysql-bin.000002
/*!50530 SET @@SESSION.PSEUDO_SLAVE_MODE=1*/;
/*!50003 SET @OLD_COMPLETION_TYPE=@@COMPLETION_TYPE,COMPLETION_TYPE=0*/;
DELIMITER /*!*/;
# at 4
#170412 12:05:32 server id 1 end_log_pos 123 CRC32 0x6d8c069c Start: binlog v 4, server v 5.7.17-log created 170412 12:05:32 at startup
# Warning: this binlog is either in use or was not closed properly.
ROLLBACK/*!*/;
BINLOG '
jKftWA8BAAAAdwAAAHsAAAABAAQANS43LjE3LWxvZwAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAACMp+1YEzgNAAgAEgAEBAQEEgAAXwAEGggAAAAICAgCAAAACgoKKioAEjQA
AZwGjG0=
'/*!*/;
# at 123
#170412 12:05:32 server id 1 end_log_pos 154 CRC32 0x17f50164 Previous-GTIDs
# [empty]
# at 154
#170412 12:05:59 server id 1 end_log_pos 219 CRC32 0x4ba5a976 Anonymous_GTID last_committed=0 sequence_number=1
SET @@SESSION.GTID_NEXT= 'ANONYMOUS'/*!*/;
# at 219
#170412 12:05:59 server id 1 end_log_pos 310 CRC32 0x5b66ae13 Query thread_id=3 exec_time=0 error_code=0
SET TIMESTAMP=1491969959/*!*/;
SET @@session.pseudo_thread_id=3/*!*/;
SET @@session.foreign_key_checks=1, @@session.sql_auto_is_null=0, @@session.unique_checks=1, @@session.autocommit=1/*!*/;
SET @@session.sql_mode=1436549152/*!*/;
SET @@session.auto_increment_increment=1, @@session.auto_increment_offset=1/*!*/;
/*!C utf8 *//*!*/;
SET @@session.character_set_client=33,@@session.collation_connection=33,@@session.collation_server=8/*!*/;
SET @@session.lc_time_names=0/*!*/;
SET @@session.collation_database=DEFAULT/*!*/;
CREATE DATABASE db1
/*!*/;
# at 310
#170412 12:06:23 server id 1 end_log_pos 375 CRC32 0x2967cc28 Anonymous_GTID last_committed=1 sequence_number=2
SET @@SESSION.GTID_NEXT= 'ANONYMOUS'/*!*/;
# at 375
#170412 12:06:23 server id 1 end_log_pos 502 CRC32 0x5de09aae Query thread_id=3 exec_time=0 error_code=0
use `db1`/*!*/;
SET TIMESTAMP=1491969983/*!*/;
CREATE TABLE tb1(
id int(4) NOT NULL,name varchar(24)
)
/*!*/;
# at 502
#170412 12:06:55 server id 1 end_log_pos 567 CRC32 0x0b8cd418 Anonymous_GTID last_committed=2 sequence_number=3
SET @@SESSION.GTID_NEXT= 'ANONYMOUS'/*!*/;
# at 567
#170412 12:06:55 server id 1 end_log_pos 644 CRC32 0x7e8f2fa0 Query thread_id=3 exec_time=0 error_code=0
SET TIMESTAMP=1491970015/*!*/;
BEGIN
/*!*/;
# at 644
#170412 12:06:55 server id 1 end_log_pos 772 CRC32 0x4e3f728e Query thread_id=3 exec_time=0 error_code=0 //插入表记录的起始时间点
SET TIMESTAMP=1491970015/*!*/;
INSERT INTO tb1 VALUES(1,'Jack'),(2,'Kenthy'), (3,'Bob')
/*!*/;
# at 772
#170412 12:06:55 server id 1 end_log_pos 803 CRC32 0x6138b21f Xid = 10
//确认事务的时间点
COMMIT/*!*/;
# at 803
#170412 12:07:24 server id 1 end_log_pos 868 CRC32 0xbef3f472 Anonymous_GTID last_committed=3 sequence_number=4
SET @@SESSION.GTID_NEXT= 'ANONYMOUS'/*!*/;
# at 868
#170412 12:07:24 server id 1 end_log_pos 945 CRC32 0x5684e92c Query thread_id=3 exec_time=0 error_code=0
SET TIMESTAMP=1491970044/*!*/;
BEGIN
/*!*/;
# at 945
#170412 12:07:24 server id 1 end_log_pos 1032 CRC32 0x4c1c75fc Query thread_id=3 exec_time=0 error_code=0 //删除表记录的时间点
SET TIMESTAMP=1491970044/*!*/;
DELETE FROM tb1
/*!*/;
# at 1032
#170412 12:07:24 server id 1 end_log_pos 1063 CRC32 0xccf549b2 Xid = 12
COMMIT/*!*/;
SET @@SESSION.GTID_NEXT= 'AUTOMATIC' /* added by mysqlbinlog */ /*!*/;
DELIMITER ;
# End of log file
/*!50003 SET COMPLETION_TYPE=@OLD_COMPLETION_TYPE*/;
/*!50530 SET @@SESSION.PSEUDO_SLAVE_MODE=0*/;
2） 执行指定Pos节点范围内的sql命令恢复数据
根据上述日志分析，只要恢复从2014.01.12 20:12:14到2014.01.12 20:13:50之间的操作即可。可通过mysqlbinlog指定时间范围输出，结合管道交给msyql命令执行导入重做：
[root@dbsvr1 ~]# mysqlbinlog
--start-datetime="2017-04-12 12:06:55"
--stop-datetime="2017-04-12 12:07:23"
/var/lib/mysql/mysql-bin.000002 | mysql -u root -p
Enter password: //验证口令
3）确认恢复结果
mysql> SELECT * FROM db1.tb1;
+----+--------+
| id | name |
+----+--------+
| 1 | Jack |
| 2 | Kenthy |
| 3 | Bob |
+----+--------+
3 rows in set (0.00 sec)
二、XtraBackup备份工具
目标：
- 安装XtraBackup软件包。
- 使用XtraBackup执行完整备份、增量备份。
- 准备数据恢复目录。
步骤：
步骤一：安装XtraBackup软件包
1）了解软件包描述信息
[root@dbsvr1 pub]# rpm -qpi percona-xtrabackup-24-2.4.6-2.el7.x86_64.rpm
Name : percona-xtrabackup-24
Version : 2.4.6
Release : 2.el7
Architecture: x86_64
Install Date: (not installed)
Group : Applications/Databases
Size : 32416340
License : GPLv2
Signature : DSA/SHA1, 2017年02月27日 星期一 20时28分17秒, Key ID 1c4cbdcdcd2efd2a
Source RPM : percona-xtrabackup-24-2.4.6-2.el7.src.rpm
Build Date : 2017年02月27日 星期一 20时27分21秒
Build Host : vps-centos7-x64-01.ci.percona.com
Relocations : (not relocatable)
URL : http://www.percona.com/software/percona-xtrabackup
Summary : XtraBackup online backup for MySQL / InnoDB
Description :
Percona XtraBackup is OpenSource online (non-blockable) backup solution for InnoDB and XtraDB engines
2）安装依赖包perl-DBD-MySQL perl-Digest-MD5 libev
使用RHEL 7自带的即可，yum方式安装：
[root@dbsvr1 pub]# yum -y install perl-DBD-MySQL perl-Digest-MD5
libev使用网上找的rpm包 libev-4.15-1.el6.rf.x86_64.rpm //该包由讲师提供
[root@dbsvr1 pub]#rpm –ivh libev-4.15-1.el6.rf.x86_64.rpm
如果未安装这些依赖包，则直接安装percona-xtrabackup时会报错：
[root@dbsvr1 pub]# rpm -ivh percona-xtrabackup-24-2.4.6-2.el7.x86_64.rpm
libev.so.4()(64bit) 被 percona-xtrabackup-24-2.4.6-2.el7.x86_64 需要
perl(DBD::mysql) 被 percona-xtrabackup-24-2.4.6-2.el7.x86_64 需要
perl(Digest::MD5) 被 percona-xtrabackup-24-2.4.6-2.el7.x86_64 需要
3）安装percona-xtrabackup
[root@dbsvr1 pub]#rpm -ivh percona-xtrabackup-*.rpm
警告：percona-xtrabackup-24-2.4.6-2.el7.x86_64.rpm: 头V4 DSA/SHA1 Signature, 密钥 ID cd2efd2a: NOKEY
准备中... ################################# [100%]
正在升级/安装...
1:percona-xtrabackup-24-2.4.6-2.el7################################# [ 33%]
2:percona-xtrabackup-test-24-2.4.6-################################# [ 67%]
3:percona-xtrabackup-24-debuginfo-2################################# [100%]
4）确认安装的主要程序/脚本
[root@dbsvr1 pub]# rpm -ql percona-xtrabackup-24-2.4.6-2.el7.x86_64
/usr/bin/innobackupex
/usr/bin/xbcloud
/usr/bin/xbcloud_osenv
/usr/bin/xbcrypt
/usr/bin/xbstream
/usr/bin/xtrabackup
/usr/share/doc/percona-xtrabackup-24-2.4.6
/usr/share/doc/percona-xtrabackup-24-2.4.6/COPYING
/usr/share/man/man1/innobackupex.1.gz
/usr/share/man/man1/xbcrypt.1.gz
/usr/share/man/man1/xbstream.1.gz
/usr/share/man/man1/xtrabackup.1.gz
步骤二：使用XtraBackup执行数据库备
--host 主机名
--port 3306
--user 用户名
--password 密码
--databases="库名"
--databases="库1 库2"
--databases="库.表"
--no-timestamp 不用日期命名备份文件存储的子目录，使用备份的数据库名做备份目录名
--no-timestmap 不使用日期命名备份目录名
1）做一个完整备份
默认情况下，备份文件存储的子目录会用日期命名，
innobackupex作为客户端工具，以mysql协议连入mysqld，将数据备份到/backup文件夹：
[root@dbsvr1 ~]# innobackupex --user=root --password=123456 /backup/mysql –no-timestamp
170425 11:05:44 innobackupex: Starting the backup operation
IMPORTANT: Please check that the backup run completes successfully.
At the end of a successful backup run innobackupex
prints "completed OK!".
Unrecognized character x01; marked by <-- HERE after <-- HERE near column 1 at - line 1374.
170425 11:05:45 Connecting to MySQL server host: localhost, user: root, password: set, port: not set, socket: not set
Using server version 5.7.17
innobackupex version 2.4.6 based on MySQL server 5.7.13 Linux (x86_64) (revision id: 8ec05b7)
xtrabackup: uses posix_fadvise().
xtrabackup: cd to /var/lib/mysql
xtrabackup: open files limit requested 0, set to 1024
xtrabackup: using the following InnoDB configuration:
xtrabackup: innodb_data_home_dir = .
xtrabackup: innodb_data_file_path = ibdata1:12M:autoextend
xtrabackup: innodb_log_group_home_dir = ./
xtrabackup: innodb_log_files_in_group = 2
xtrabackup: innodb_log_file_size = 50331648
InnoDB: Number of pools: 1
170425 11:05:45 >> log scanned up to (2543893)
xtrabackup: Generating a list of tablespaces
InnoDB: Allocated tablespace ID 2 for mysql/plugin, old maximum was 0
170425 11:05:45 [01] Copying ./ibdata1 to /backup/ibdata1
170425 11:05:45 [01] ...done
170425 11:05:46 [01] Copying ./mysql/plugin.ibd to /backup/mysql/plugin.ibd
170425 11:05:46 [01] ...done
170425 11:05:46 [01] Copying ./mysql/servers.ibd to /backup/mysql/servers.ibd
170425 11:05:46 [01] ...done
170425 11:05:46 [01] Copying ./mysql/help_topic.ibd to /backup/mysql/help_topic.ibd
170425 11:05:46 [01] ...done
170425 11:05:46 >> log scanned up to (2543893)
.. ..
170425 11:06:00 [01] Copying ./sys/x@0024waits_global_by_latency.frm to /backup/sys/x@0024waits_global_by_latency.frm
170425 11:06:00 [01] ...done
170425 11:06:00 [01] Copying ./sys/session_ssl_status.frm to /backup/sys/session_ssl_status.frm
170425 11:06:00 [01] ...done
170425 11:06:00 [01] Copying ./db1/db.opt to /backup/db1/db.opt
170425 11:06:00 [01] ...done
170425 11:06:00 [01] Copying ./db1/tb1.frm to /backup/db1/tb1.frm
170425 11:06:00 [01] ...done
170425 11:06:00 Finished backing up non-InnoDB tables and files
170425 11:06:00 Executing FLUSH NO_WRITE_TO_BINLOG ENGINE LOGS...
xtrabackup: The latest check point (for incremental): '2543884'
xtrabackup: Stopping log copying thread.
.170425 11:06:00 >> log scanned up to (2543893)
170425 11:06:00 Executing UNLOCK TABLES
170425 11:06:00 All tables unlocked
170425 11:06:00 [00] Copying ib_buffer_pool to /backup/ib_buffer_pool
170425 11:06:00 [00] ...done
170425 11:06:00 Backup created in directory '/backup/'
170425 11:06:00 [00] Writing backup-my.cnf
170425 11:06:00 [00] ...done
170425 11:06:00 [00] Writing xtrabackup_info
170425 11:06:00 [00] ...done
xtrabackup: Transaction log of lsn (2543884) to (2543893) was copied.
170425 11:06:01 completed OK
确认备份好的文件数据：
[root@dbsvr1 ~]#ls /backup/mysql
backup-my.cnf ib_buffer_pool mysql sys xtrabackup_info
db1 ibdata1 performance_schema xtrabackup_checkpoints xtrabackup_logfile
2）做一个增量备份（基于前一步的完整备份）
随意做一些新增或更改库表的操作，比如在db1库中新建一个mytb的表：
mysql> USE db1;
Database changed
mysql> CREATE TABLE mytb(id int(4), name varchar(24));
Query OK, 0 rows affected (0.38 sec)
mysql> INSERT INTO tb1 VALUES
-> (1,'bon'),
-> (2,'bo'),
Query OK, 2 rows affected (0.12 sec)
Records: 2 Duplicates: 0 Warnings: 0
mysql> SELECT * FROM tb1;
+------+------+
| id | name |
+------+------+
| 1 | bob |
| 2 | bo |
+------+------+
2 rows in set (0.00 sec)
以前一次保存到/backup的完整备份为基础，做一个增量备份，保存到/incr01/，指定增量备份参照的基本目录（完整备份目录）需要用到选项--incremental-basedir。相关操作如下：
[root@dbsvr1 ~]# innobackupex --user=root --password=123456 --incremental /incr01 --incremental-basedir=/backup/mysql --no-timestamp
170425 11:30:14 innobackupex: Starting the backup operation
IMPORTANT: Please check that the backup run completes successfully.
At the end of a successful backup run innobackupex
prints "completed OK!".
Unrecognized character x01; marked by <-- HERE after <-- HERE near column 1 at - line 1374.
170425 11:30:14 Connecting to MySQL server host: localhost, user: root, password: set, port: not set, socket: not set
Using server version 5.7.17
innobackupex version 2.4.6 based on MySQL server 5.7.13 Linux (x86_64) (revision id: 8ec05b7)
incremental backup from 2543884 is enabled.
xtrabackup: uses posix_fadvise().
xtrabackup: cd to /var/lib/mysql
xtrabackup: open files limit requested 0, set to 1024
xtrabackup: using the following InnoDB configuration:
xtrabackup: innodb_data_home_dir = .
xtrabackup: innodb_data_file_path = ibdata1:12M:autoextend
xtrabackup: innodb_log_group_home_dir = ./
xtrabackup: innodb_log_files_in_group = 2
xtrabackup: innodb_log_file_size = 50331648
InnoDB: Number of pools: 1
170425 11:30:14 >> log scanned up to (2549933)
xtrabackup: Generating a list of tablespaces
InnoDB: Allocated tablespace ID 2 for mysql/plugin, old maximum was 0
xtrabackup: using the full scan for incremental backup
170425 11:30:15 [01] Copying ./ibdata1 to /incr01/ibdata1.delta
170425 11:30:15 [01] ...done
170425 11:30:15 >> log scanned up to (2549933)
170425 11:30:15 [01] Copying ./mysql/plugin.ibd to /incr01/mysql/plugin.ibd.delta
170425 11:30:15 [01] ...done
... ...
170425 11:30:35 Executing UNLOCK TABLES
170425 11:30:35 All tables unlocked
170425 11:30:35 [00] Copying ib_buffer_pool to /incr01/ib_buffer_pool
170425 11:30:35 [00] ...done
170425 11:30:35 Backup created in directory '/incr01/'
170425 11:30:35 [00] Writing backup-my.cnf
170425 11:30:35 [00] ...done
170425 11:30:35 [00] Writing xtrabackup_info
170425 11:30:35 [00] ...done
xtrabackup: Transaction log of lsn (2549924) to (2549933) was copied.
170425 11:30:35 completed OK!
确认备份好的文件数据：
[root@dbsvr1 ~]# ls /incr01/
backup-my.cnf ib_buffer_pool ibdata1.meta performance_schema xtrabackup_checkpoints xtrabackup_logfile
db1 ibdata1.delta mysql sys
对比完整备份、增量备份的大小：
[root@dbsvr1 ~]# du -sh /backup/mysql /incr01/
142M /backup/mysql //完整备份的大小
3.5M /incr01/ //增量备份的大小
步骤三：准备用于恢复的数据库目录
通过XtraBackup工具备份的数据库目录，若要恢复到另一个MySQL服务器，需要先做一个“--apply-log --redo-only ”的准备操作。
1）准备恢复“完整备份”
完成准备以后，最终/backup可用来重建MySQL服务器。这种情况下，需要先做一个“--apply-log --redo-only ”的准备操作，以确保数据一致性：
[root@dbsvr1 ~]#innobackupex --user=root --password=123456 --apply-log --redo-only /backup/mysql //对备份数据库的位置做日志回读操作
170425 11:42:19 innobackupex: Starting the apply-log operation
IMPORTANT: Please check that the apply-log run completes successfully.
At the end of a successful apply-log run innobackupex
prints "completed OK!".
innobackupex version 2.4.6 based on MySQL server 5.7.13 Linux (x86_64) (revision id: 8ec05b7)
xtrabackup: cd to /backup/mysql
xtrabackup: This target seems to be already prepared.
InnoDB: Number of pools: 1
xtrabackup: notice: xtrabackup_logfile was already used to '--prepare'.
xtrabackup: using the following InnoDB configuration for recovery:
xtrabackup: innodb_data_home_dir = .
xtrabackup: innodb_data_file_path = ibdata1:12M:autoextend
xtrabackup: innodb_log_group_home_dir = .
xtrabackup: innodb_log_files_in_group = 2
xtrabackup: innodb_log_file_size = 50331648
xtrabackup: using the following InnoDB configuration for recovery:
xtrabackup: innodb_data_home_dir = .
xtrabackup: innodb_data_file_path = ibdata1:12M:autoextend
xtrabackup: innodb_log_group_home_dir = .
xtrabackup: innodb_log_files_in_group = 2
xtrabackup: innodb_log_file_size = 50331648
xtrabackup: Starting InnoDB instance for recovery.
xtrabackup: Using 104857600 bytes for buffer pool (set by --use-memory parameter)
InnoDB: PUNCH HOLE support available
InnoDB: Mutexes and rw_locks use GCC atomic builtins
InnoDB: Uses event mutexes
InnoDB: GCC builtin __atomic_thread_fence() is used for memory barrier
InnoDB: Compressed tables use zlib 1.2.7
InnoDB: Number of pools: 1
InnoDB: Not using CPU crc32 instructions
InnoDB: Initializing buffer pool, total size = 100M, instances = 1, chunk size = 100M
InnoDB: Completed initialization of buffer pool
InnoDB: page_cleaner coordinator priority: -20
InnoDB: Highest supported file format is Barracuda.
xtrabackup: starting shutdown with innodb_fast_shutdown = 1
InnoDB: Starting shutdown...
InnoDB: Shutdown completed; log sequence number 2544177
InnoDB: Number of pools: 1
170425 11:42:20 completed OK!
准备恢复“增量备份”
[root@dbsvr1 ~]#innobackupex --user=root --password=123456 --apply-log --redo-only /backup/mysql --incremental-dir=/incr01
170425 11:42:55 innobackupex: Starting the apply-log operation
IMPORTANT: Please check that the apply-log run completes successfully.
At the end of a successful apply-log run innobackupex
prints "completed OK!".
innobackupex version 2.4.6 based on MySQL server 5.7.13 Linux (x86_64) (revision id: 8ec05b7)
incremental backup from 2543884 is enabled.
xtrabackup: cd to /backup/
xtrabackup: This target seems to be already prepared with --apply-log-only.
InnoDB: Number of pools: 1
xtrabackup: xtrabackup_logfile detected: size=8388608, start_lsn=(2549924)
xtrabackup: using the following InnoDB configuration for recovery:
xtrabackup: innodb_data_home_dir = .
xtrabackup: innodb_data_file_path = ibdata1:12M:autoextend
xtrabackup: innodb_log_group_home_dir = /incr01/
xtrabackup: innodb_log_files_in_group = 1
xtrabackup: innodb_log_file_size = 8388608
xtrabackup: Generating a list of tablespaces
InnoDB: Allocated tablespace ID 2 for mysql/plugin, old maximum was 0
xtrabackup: page size for /incr01//ibdata1.delta is 16384 bytes
Applying /incr01//ibdata1.delta to ./ibdata1...
... ...
170425 11:43:09 [01] Copying /incr01/performance_schema/global_status.frm to ./performance_schema/global_status.frm
170425 11:43:09 [01] ...done
170425 11:43:09 [01] Copying /incr01/performance_schema/session_status.frm to ./performance_schema/session_status.frm
170425 11:43:09 [01] ...done
170425 11:43:09 [00] Copying /incr01//xtrabackup_info to ./xtrabackup_info
170425 11:43:09 [00] ...done
170425 11:43:10 completed OK!
2）关闭mysql服务，并将/var/lib/mysql/下的文件删除，假设数据被删除。
[root@dbsvr1 ~]#systemctl stop mysqld
[root@dbsvr1 ~]#rm -rf /var/lib/mysql
3）恢复“完整备份+增量备份”
完成准备以后，最终仍然是/backup用来重建MySQL服务器，但这种情况下需提前合并相关增量备份的数据
[root@dbsvr1 ~]# mkdir /var/lib/mysql //注意恢复数据库前，重建/var/lib/mysql数据库目录，因为刚才为了做假设将目录删掉了；后面测试发现这个目录在innobackupex命令下会自动创建，所以这步骤可以没有
[root@dbsvr1 ~]# chown -R mysql:mysql /var/lib/mysql //创建的目录需要归档为mysql用户和mysql组，否则启动会报错，保证mysql用户对此目录的权限；如果上面没创建目录这步骤可以不要，后面等innobackupex恢复命令执行后（自动创建这个目录），再归档
[root@dbsvr1 ~]# innobackupex --user=root --password=123456 --copy-back /backup/mysql //将backup/mysql里面的数据恢复回来
... ...
170425 11:51:39 [01] Copying ./performance_schema/global_status.frm to /var/lib/mysql/performance_schema/glo.frm
170425 11:51:39 [01] ...done
170425 11:51:39 [01] Copying ./performance_schema/session_status.frm to /var/lib/mysql/performance_schema/seus.frm
170425 11:51:39 [01] ...done
170425 11:51:39 [01] Copying ./ib_buffer_pool to /var/lib/mysql/ib_buffer_pool
170425 11:51:39 [01] ...done
170425 11:51:39 [01] Copying ./ibtmp1 to /var/lib/mysql/ibtmp1
170425 11:51:39 [01] ...done
170425 11:51:39 [01] Copying ./xtrabackup_info to /var/lib/mysql/xtrabackup_info
170425 11:51:39 [01] ...done
170425 11:51:39 completed OK!
4）修改/var/lib/mysql/下文件属主与属组,查看数据：
恢复后，/var/lib/mysql下文件属组与属主皆为root，需要更改为mysql
[root@dbsvr1 ~]#chown -R mysql:mysql /var/lib/mysql //chown -R 递归归属文件属性，这步骤最好放在第三步骤操作
[root@dbsvr1 ~]#systemctl start mysqld.service
[root@dbsvr1 ~]#mysql -uroot -p12345678 -e "select * from db1.tb1"
mysql: [Warning] Using a password on the command line interface can be insecure.
+------+------+
| id | name |
+------+------+
| 1 | bob |
| 2 | bo |
+------+------+
三、MySQL AB复制
目标：
- 配置2台MySQL服务器，实现 主-->从 同步。
- 其中Master服务器允许SQL查询、写入，Slave服务器只允许SQL查询
方案：
使用2台RHEL 6虚拟机，如下图所示。其中192.168.4.10是MySQL主服务器，负责提供同步源；另一台192.168.4.20作为MySQL从服务器， 通过调取主服务器上的binlog日志，在本地重做对应的库、表，实现与主服务器的AB复制（同步）。
提前为两台MySQL服务器安装好MySQL-server、MySQL-Client软件包，并为数据库用户root修改密码；Linux客户机上则只需安装MySQL-Client软件包即可。
步骤：
步骤一：初始化现有库
为了在启用binlog日志及同步之前保持主、从库的一致性，建议进行初始化——备份主服务器上现有的库，然后导入到从服务器上。
当现有库、表都采用MyISAM引擎时，可执行离线备份、恢复，这样更有效率；否则，可通过mysqldump等工具来实现库的导出、导入。
1）备份MySQL Master（192.168.4.10）上现有的库
如果服务器已经启用binlog，建议对日志做一次重置，否则可忽略：
[root@dbsvr1 ~]# mysql -u root -p //可以这里-p后面紧跟密码，即可直接进入数据库
Enter password: //以数据库用户root登入
.. ..
mysql> RESET MASTER; //重置binlog日志
Query OK, 0 rows affected (0.06 sec)
mysql> quit //退出mysql> 环境
Bye
以备份mysql库、sys库为例，导出操作如下：
[root@dbsvr1 ~]# mysqldump -u root -p --all-databases > /root/mytest.sql //可以这里-p后面紧跟密码
Enter password: //验证口令
[root@dbsvr1 ~]# ls -lh /root/mytest.sql //确认备份结果
-rw-r--r--. 1 root root 777172 4月 23 12:21 /root/mytest.sql
2）在MySQL Slave（192.168.4.20）上导入备份的库
先清理目标库，避免导入时冲突。主要是采用InnoDB引擎的库，授权库mysql多采用MyISAM引擎，可不做清理。
[root@dbsvr2 ~]# mysql -u root -p
Enter password: //以数据库用户root登入
.. ..
mysql> DROP DATABASE test; //删除test库
Query OK, 0 rows affected (0.03 sec)
mysql> quit //退出mysql> 环境
Bye
使用scp工具下载备份文件：
[root@dbsvr1 ~]# scp /root/mytest.sql root@192.168.4.20:/
root@dbsvr1's password: //验证对方系统用户root的口令
mytest.sql 100% 759KB 759.0KB/s 00:00
[root@dbsvr1 ~]# ls -lh mytest.sql //确认下载结果
-rw-r--r--. 1 root root 759K 4月 23 12:22 /mytest.sql
执行导入操作：
[root@dbsvr2 ~]# mysql -u root -p < /mytest.sql
Enter password: //验证口令
导入成功后，可重新登入 mysql> 环境，确认清理的目标库已恢复： mysql> show databases;
+--------------------+
| Database |
+--------------------+
| information_schema |
| mysql |
| performance_schema |
| sys |
+--------------------+
4 rows in set (0.00 sec)
步骤二：配置MySQL Master（主服务器，192.168.4.10）
1）修改/etc/my.cnf配置，重新启动MySQL服务程序
指定服务器ID号、允许日志同步：
[root@dbsvr1 mysql]# vim /etc/my.cnf
[mysqld]
log_bin=dbsvr1-bin //启用binlog日志，并指定文件名前缀
server_id = 10 //指定服务器ID号
......
重启mysql服务：
[root@dbsvr1 ~]# systemctl restart mysqld.service
2）新建一个备份用户，授予复制权限
需要的权限为REPLICATION SLAVE，允许其从Slave服务器访问：
mysql> GRANT REPLICATION SLAVE ON *.* TO 'replicater'@'192.168.4.%' IDENTIFIED BY 'pwd123';
Query OK, 0 rows affected, 1 warning (0.09 sec)
3）检查Master服务器的同步状态
在已经初始化现有库的情况下，查看MASTER状态，记录下当前的日志文件名、偏移的位置（下面SLAVE发起复制时需要用到）：
mysql> SHOW MASTER STATUSG
*************************** 1. row ***************************
File: dbsvr1-bin.000001 //记住当前的日志文件名
Position: 154 //记住当前的位置
Binlog_Do_DB:
Binlog_Ignore_DB:
Executed_Gtid_Set:
1 row in set (0.00 sec)
步骤三：配置MySQL Slave（从服务器，192.168.4.20）
1）修改/etc/my.cnf配置，重新启动MySQL服务程序
指定服务器ID号、允许日志同步： [root@dbsvr2 ~]# vim /etc/my.cnf
[mysqld]
log_bin=dbsvr2-bin //启动SQL日志，并指定文件名前缀
server_id = 20 //指定服务器ID号，不要与Master的相同
.. ..
在生产环境中，还可以根据需要设置更详细的同步选项。比如，指定当主、从网络中断时的重试超时时间（slave-net-timeout=60 ）等，具体可参考MySQL手册。
配置完成后，重启mysql服务：
[root@dbsvr2 ~]# systemctl restart mysqld.service
通过CHANGE MASTER语句指定MASTER服务器的IP地址、同步用户名/密码、起始日志文件、偏移位置（参考MASTER上的状态输出）：
mysql> CHANGE MASTER TO MASTER_HOST='192.168.4.10',
-> MASTER_USER='replicater',
-> MASTER_PASSWORD='pwd123',
-> MASTER_LOG_FILE='dbsvr1-bin.000002', //对应Master的日志文件
-> MASTER_LOG_POS=334; //对应Master的日志偏移位置
Query OK, 0 rows affected, 2 warnings (0.12 sec)
然后执行START SLAVE（较早版本中为SLAVE START）启动复制：
mysql> START SLAVE; //启动复制
Query OK, 0 rows affected (0.00 sec)
注意：一旦启用SLAVE复制，当需要修改MASTER信息时，应先执行STOP SLAVE停止复制，然后重新修改、启动复制。
通过上述连接操作，MASTER服务器的设置信息自动存为master.info文件，以后每次MySQL服务程序时会自动调用并更新，无需重复设置。查看master.info文件的开头部分内容，可验证相关设置：
[root@dbsvr2 ~]# ls -lh /var/lib/mysql/master.info
-rw-r-----. 1 mysql mysql 132 4月 23 12:06 /var/lib/mysql/master.info
[root@dbsvr2 ~]# head /var/lib/mysql/master.info //注意这个文件是保存在slave服务器的
25
dbsvr1-bin.000001
154
192.168.4.10
replicater
pwd123
3306
60
0
2）检查Slave服务器的同步状态
通过SHOW SLAVE STATUS语句可查看从服务器状态，确认其中的IO线程、SQL线程正常运行，才能成功同步：
mysql> SHOW SLAVE STATUSG
Slave_IO_State: Waiting for master to send event
Master_Host: 192.168.4.1
Master_User: replicater
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: dbsvr1-bin.000001
Read_Master_Log_Pos: 154
Relay_Log_File: db2-relay-bin.000003
Relay_Log_Pos: 321
Relay_Master_Log_File: dbsvr1-bin.000001
Slave_IO_Running: Yes //IO线程应该已运行
Slave_SQL_Running: Yes //SQL线程应该已运行
Replicate_Do_DB:
Replicate_Ignore_DB:
Replicate_Do_Table:
Replicate_Ignore_Table:
Replicate_Wild_Do_Table:
Replicate_Wild_Ignore_Table:
Last_Errno: 0
Last_Error:
Skip_Counter: 0
Exec_Master_Log_Pos: 154
Relay_Log_Space: 2490
Until_Condition: None
Until_Log_File:
Until_Log_Pos: 0
Master_SSL_Allowed: No
Master_SSL_CA_File:
Master_SSL_CA_Path:
Master_SSL_Cert:
Master_SSL_Cipher:
Master_SSL_Key:
Seconds_Behind_Master: 0
Master_SSL_Verify_Server_Cert: No
Last_IO_Errno: 0
Last_IO_Error:
Last_SQL_Errno: 0
Last_SQL_Error:
Replicate_Ignore_Server_Ids:
Master_Server_Id: 10
Master_UUID: 2d4d8a11-27b7-11e7-ae78-52540055c180
Master_Info_File: /var/lib/mysql/master.info
SQL_Delay: 0
SQL_Remaining_Delay: NULL
Slave_SQL_Running_State: Slave has read all relay log; waiting for more updates
Master_Retry_Count: 86400
Master_Bind:
Last_IO_Error_Timestamp:
Last_SQL_Error_Timestamp:
Master_SSL_Crl:
Master_SSL_Crlpath:
Retrieved_Gtid_Set:
Executed_Gtid_Set:
Auto_Position: 0
Replicate_Rewrite_DB:
Channel_Name:
Master_TLS_Version:
1 row in set (0.00 sec)
若START SLAVE直接报错失败，请检查CHANGE MASTER相关设置是否有误，纠正后再重试；若IO线程或SQL线程有一个为“No”，则应检查服务器的错误日志，分析并排除故障后重启主从复制。
步骤四：测试主从同步效果
1）在Master上操作数据库、表、表记录
新建newdb库、newtable表，随意插入几条表记录：
mysql> CREATE DATABASE newdb; //新建库newdb
Query OK, 1 row affected (0.17 sec)
mysql> USE newdb; //切换到newdb库
Database changed
mysql> CREATE TABLE newtable(id int(4)); //新建newtable表
Query OK, 0 rows affected (0.46 sec)
mysql> INSERT INTO newtable VALUES(1234),(5678); //插入2条表记录
Query OK, 2 rows affected (0.24 sec)
Records: 2 Duplicates: 0 Warnings: 0
mysql> SELECT * FROM newtable; //确认表数据
+------+
| id |
+------+
| 1234 |
| 5678 |
+------+
2 rows in set (0.00 sec)
2）在Slave上确认自动同步的结果
直接切换到newdb库，并查询newtable表的记录，应该与Master上的一样，这才说明主从同步已经成功生效：
mysql> USE newdb; //直接切换到newdb库
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A
Database changed
mysql> SELECT * FROM newtable; //输出表记录
+------+
| id |
+------+
| 1234 |
| 5678 |
+------+
2 rows in set (0.02 sec)
3）在Master服务器上可查看Slave主机的信息
mysql> SHOW SLAVE HOSTS;
+-----------+------+------+-----------+--------------------------------------+
| Server_id | Host | Port | Master_id | Slave_UUID |
+-----------+------+------+-----------+--------------------------------------+
| 2 | | 3306 | 10 | 512cf7c1-27c4-11e7-8f4b-5254007b030b |
+-----------+------+------+-----------+--------------------------------------+
1 row in set (0.00 sec)
步骤五：将Slave服务器设为只读
一般来说，为了避免写入冲突，采用主、从复制结构时，不应该允许用户从Slave执行数据库写入操作，这样会导致双方数据的不一致性。
正因为如此，我们可以把Slave数据库限制为只读模式，这种情况下有SUPER权限的用户和SLAVE同步线程才能写入。相关验证操作及效果可参考以下过程。
1）新建一个测试用户rwuser（不能用root测试）
在Master上建立即可，会自动同步到Slave上：
mysql> GRANT all ON newdb.* TO rwuser@localhost IDENTIFIED BY '1234567';
Query OK, 0 rows affected, 1 warning (0.09 sec)
2）未启用只读前，验证从Slave写入
在Slave上以rwuser登入（不要用root哦）：
[root@dbsvr2 ~]# mysql -u rwuser -p
Enter password:
Welcome to the MySQL monitor. Commands end with ; or g.
Your MySQL connection id is 30
Server version: 5.7.17-log MySQL Community Server (GPL)
Copyright (c) 2000, 2016, Oracle and/or its affiliates. All rights reserved.
Oracle is a registered trademark of Oracle Corporation and/or its
affiliates. Other names may be trademarks of their respective
owners.
Type 'help;' or 'h' for help. Type 'c' to clear the current input statement.
mysql>
然后向newdb库中新建一个booker表：
mysql> USE newdb; //切换到newdb库
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A
Database changed
mysql> CREATE TABLE booker(id int(12)); //成功创建booker表
Query OK, 0 rows affected (0.64 sec)
在Slave上可看到新建的booker表：
mysql> SHOW TABLES;
+-----------------+
| Tables_in_newdb |
+-----------------+
| booker |
| newtable |
+-----------------+
2 rows in set (0.00 sec)
但是在Master上却看不到，导致主、从上的newdb出现不一致：
mysql> USE newdb;
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A
Database changed
mysql> SHOW TABLES; //看不到Slave上新建的表
+-----------------+
| Tables_in_newdb |
+-----------------+
| newtable |
+-----------------+
1 row in set (0.00 sec)
完成上述验证后，在Slave上删除booker表，确保双方一致：
mysql> DROP TABLE booker;
Query OK, 0 rows affected (0.27 sec)
3）修改/etc/my.cnf文件，重载配置
[root@dbsvr2 ~]# vim /etc/my.cnf
[mysqld]
.. ..
read_only=1 //启动只读模式
[root@dbsvr2 ~]# systemctl restart mysqld.service //重启服务
4）再次在Slave上验证数据库写入操作
仍然是以rwuser登入（不要用root哦）来验证，当尝试创建新表时会被拒绝：
mysql> USE newdb; //切换到newdb库
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A
Database changed
mysql> CREATE TABLE booker(id int(12)); //新建表的写入操作失败
ERROR 1290 (HY000): The MySQL server is running with the --read-only option so it cannot execute this statement
mysql> DROP TABLE mytable; //删除表的写入操作一样会失败
ERROR 1290 (HY000): The MySQL server is running with the --read-only option so it cannot execute this statement
附加：
一、设置binlog日志，修改配置文件后，启动mysqld服务报错，排错思路
1> 检查SELinux是否设置为Permissing，如果没有就改为Permissing，可以直接设置setenforce 0【临时Permissive】，也可以修改/etc/selinux/config配置文件，设置selinux永久Permissive
2> 检查/etc/my.cnf配置文件里binlog日志文件log_bin=/tmp/logdir/plj的修改路径/tmp/logdir/plj有没有创建
3> 创建了路径目录/tmp/logdir/plj后，需要chown mysql /tmp/logdir 将文件目录归属给mysql用户，注意：备份的日志文件都在/tmp/logdir/下面，而不是/tmp/logdir/plj下面
4> 检查httpd服务是否开启，有授权给网段用户可以登录的话，必须启动httpd才能正常启动mysqld
二、mysql数据备份恢复，启动mysqld服务报错，排错思路
1> mysql在设置备份服务后，备份完成后，重启mysqld服务失败，检查/var/lib/mysql目录是否归档为mysql用户和mysql组。
全备份和增量备份思路：
备份：
# innobackupex --user=root --password=123456 /backup/mysql –no-timestamp //全备份
# innobackupex --user=root --password=123456 --incremental /incr01 --incremental-basedir=/backup/mysql --no-timestamp //增量备份
# du -sh /backup/ /incr01/ //查看全备份和增量备份的文件大小
恢复：
# innobackupex --user=root --password=123456 --apply-log --redo-only /backup/mysql //全备份的恢复
# innobackupex --user=root --password=123456 --apply-log --redo-only /backup/mysql --incremental-dir=/incr01 //增量备份的恢复
模拟数据库误删除操作：
# systemctl stop mysqld
# rm -rf /var/lib/mysql
重建数据库目录并归档：
# mkdir /var/lib/mysql //重建/var/lib/mysql数据库目录，这步骤可以不要，下面innobackupex会自动创建
# chown -R mysql:mysql /var/lib/mysql //保证mysql用户对此目录的权限，上面若没有创建目录，这步骤可以放到下面
# innobackupex --user=root --password=123456 --copy-back /backup/mysql //将backup/mysql里面的数据恢复回来
# chown -R mysql:mysql /var/lib/mysql //保证mysql用户对此目录的权限，上面若有这步骤可省略
# systemctl start mysqld.service //重启数据库服务，登录验证
附加：
摘自：
http://www.linuxidc.com/Linux/2014-02/96945.htm
MySQL主从同步故障－Slave_SQL_Running: No
故障现象：
进入slave服务器，运行：
mysql> show slave statusG
.......
Relay_Log_File: localhost-relay-bin.000535
Relay_Log_Pos: 21795072
Relay_Master_Log_File: localhost-bin.000094
Slave_IO_Running: Yes
Slave_SQL_Running: No
Replicate_Do_DB:
Replicate_Ignore_DB:
......
解决办法一、
Slave_SQL_Running: No
1.程序可能在slave上进行了写操作
2.也可能是slave机器重起后，事务回滚造成的.
一般是事务回滚造成的：
解决办法：
mysql> slave stop;
mysql> set GLOBAL SQL_SLAVE_SKIP_COUNTER=1;
mysql> slave start;
解决办法二、
首先停掉Slave服务：slave stop
到主服务器上查看主机状态：
记录File和Position对应的值
进入master
mysql> show master status;
+----------------------+----------+--------------+------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+----------------------+----------+--------------+------------------+
| localhost-bin.000094 | 33622483 | | |
+----------------------+----------+--------------+------------------+
1 row in set (0.00 sec)
然后到slave服务器上执行手动同步：
mysql> change master to
> master_host='master_ip',
> master_user='user',
> master_password='pwd',
> master_port=3306,
> master_log_file=localhost-bin.000094',
> master_log_pos=33622483 ;
1 row in set (0.00 sec)
mysql> slave start;
1 row in set (0.00 sec)
mysql> show slave statusG
*************************** 1. row ***************************
........
Master_Log_File: localhost-bin.000094
Read_Master_Log_Pos: 33768775
Relay_Log_File: localhost-relay-bin.000537
Relay_Log_Pos: 1094034
Relay_Master_Log_File: localhost-bin.000094
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
Replicate_Do_DB:
手动同步需要停止master的写操作！
查看mysql主从配置的状态及修正 slave不启动问题
1、查看master的状态
show master status; //Position不应该为0
show processlist;
//state状态应该为Has sent all binlog to slave; waiting for binlog to be updated
2、查看slave状态
show slave status;
//Slave_IO_Running 与 Slave_SQL_Running 状态都要为Yes
show processlist;
//应该有两行state值为：
Has read all relay log; waiting for the slave I/O thread to update it
Waiting for master to send event
3、错误日志
MySQL安装目录 /usr/local/mysql
MySQL日志目录 /usr/local/mysql/data/ 形如，Hostname.err
4、Change master to
如果从库的Slave未启动，Slave_IO_Running为NO。
可能是主库是的master的信息有变化，
查看主库show master status;
记录下File,Position字段，假设为‘mysql-bin.000004’,98;
在从库执行：
mysql>stop slave;
mysql>change master to master_log_file='mysql-bin.000004',master_log_pos=98;
mysql>start slave;
5、SET global sql_slave_skip_counter=n;
如果从库的slave_sql_running为NO。
Err文件中记录：
Slave:Error "Duplicate entry '1' for key 1" on query.....
可能是master未向slave同步成功，但slave中已经有了记录。造成的冲突可以在从库上执行
set global sql_slave_skip_counter=n;
跳过几步。再restart slave就可以了。
6、同步错误处理
发现mysql slave服务器经常因为一些特殊字符或者符号产生的更新语句报错，整个同步也会因此而卡在那，最初的办法只是手动去出错的机器执行下面三条SQL语句，跳过错误即可。
mysql>slave stop;
mysql>set GLOBAL SQL_SLAVE_SKIP_COUNTER=1;
mysql>slave start;
PS：本人多次遇到从数据库的同步进程自动停掉的问题，有时简单通过slave stop,slave start即可解决。有时slave start启动后又会自动停掉，这时使用 change master重设主数据库信息的方式解决了问题。
说明：
Slave_IO_Running：连接到主库，并读取主库的日志到本地，生成本地日志文件
Slave_SQL_Running:读取本地日志文件，并执行日志里的SQL命令。