xtrabackup备份和恢复MySQL

xtrabackup有两个主要的工具:innobackupexxtrabackupxtrabackup只能备份InnoDBXtraDB数据表,innobackupex封装了xtrabackup,可以备份MyISAM数据表。

MySQL数据库本身提供的工具并不支持真正的增量备份,二进制日志恢复是point-in-time(时间点)的恢复而不是增量备份。Xtrabackup工具支持对InnoDB存储引擎的增量备份,工作原理如下:

1.首先完成一个完全备份,并记录下此时检查点的LSN(Log Sequence Number)

2.在进程增量备份时,比较表空间中每个页的LSN是否大于上次备份时的LSN,如果是,则备份该页,同时记录当前检查点的LSN

 

1.下载xtrabackup(下载二进制版本免安装)

# wget http://www.percona.com/redir/downloads/XtraBackup/XtraBackup-1.5/Linux/binary/x86_64/xtrabackup-1.5.tar.gz

# tar zxvf xtrabackup-1.5.tar.gz -C /usr/local/

源码安装过程具体参见源码目录下BUILD.txt文件。

 

2.xtrabackup选项注释如下:

–print-defaults 显示默认选项。Xtrabackup默认情况会去读my.cnf文件,读取顺序是/etc/my.cnf /etc/mysql/my.cnf /usr/local/etc/my.cnf ~/.my.cnf

–no-defaults 忽略任何my.cnf文件选项

–defaults-file=# 读取指定的my.cnf文件

–defaults-extra-file=# 再读取另外一个文件

–target-dir=name 目的目录,默认目录在./xtrabackup_backupfiles/,相对于datadir目录

–backup 备份

–stats 计算datadir目录统计信息

–prepare backup恢复

–export 在恢复时,创建文件导入到另一个数据库

–apply-log-only 在恢复时,停止恢复进程不进行LSN只使用log

–print-param 打印出参数

–use-memory=# buffer_pool_size

–suspend-at-end 在备份时,创建xtrabackup_suspended文件,直到备份完成后删掉

–throttle=# 限制IO总数

–log-stream 记录标准输出信息xtrabackup_logfile

–extra-lsndir=name 仅适用于backup,保存另一份xtrabackup_checkpoints文件

–incremental-lsn=name 仅适用于backup,增量备份

–incremental-basedir=name 仅适用于backup,增量备份目录

–incremental-dir=name 仅适用于prepare,恢复指定目录下的.delta文件和日志文件

–tables=name 过滤某些表

–tables_file=name 过滤database.table列表文件

–create-ib-logfile

-h, –datadir=name datadir目录

-t, –tmpdir=name tmpdir目录

–parallel=# 默认为1.传输数据文件的并行线程数。没有任何流模式的影响

–innodb_* 有关innodb参数

 

3.innobackupex-1.5.1选项注释如下:

innobackupex-1.5.1 [–sleep=MS] [–compress[=LEVEL]] [–include=REGEXP] [–user=NAME]

           [–password=WORD] [–port=PORT] [–socket=SOCKET] [–no-timestamp]

           [–ibbackup=IBBACKUP-BINARY] [–slave-info] [–stream=tar]

           [–scpopt=OPTIONS-FOR-SCP]

           [–defaults-file=MY.CNF]

           [–databases=LIST] [–remote-host=HOSTNAME] [–no-lock] BACKUP-ROOT-DIR

对数据库进行热备。默认情况下以当前时间戳创建备份根目录。上述命令完整备份所有MyISAMInnoDB引擎表以及索引。包含.frm,.MRG, .MYD, .MYI, .TRG, .TRN, .ARM, .ARZ, .opt, .parInnoDB数据和日志文件。使用mysql客户端程序连接mysql服务器,运行ibbackup(InnoDB热备程序)作为一个子进程。

 

innobackupex-1.5.1 –apply-log [–use-memory=MB] [–uncompress] [–defaults-file=MY.CNF]

           [–export] [–redo-only] [–ibbackup=IBBACKUP-BINARY] BACKUP-DIR

从备份恢复,运行ibbackup子进程,但是不能连接到数据库。按照BACKUP-DIR/backup-my.cnf定义,使用BACKUP-DIR/ibbackup_logfile恢复InnoDB数据文件和创建新的InnoDB日志文件。

 

innobackupex-1.5.1 –copy-back [–defaults-file=MY.CNF] BACKUP-DIR

从备份目录拷贝数据,索引和日志文件到原datadir目录。

 

–defaults-file=[MY.CNF] 该选项传递给xtrabackup子进程,从指定文件读取缺省选项

–apply-log 从备份恢复。

–redo-only 该选项强制跳过rollback阶段,只进行redo。这是有必要使用的,如果备份后,要使用增量改变的。

–copy-back 从备份目录拷贝数据和索引文件到datadir目录

–remote-host=HOSTNAME 备份到远程主机上,使用ssh

–stream=[tar|cpio(not implemented)] 指定备份标准输出格式

–tmpdir=DIRECTORY 默认与tmpdir相同。使用—remote-host—stream参数后,传输日志文件将存放在临时目录下

–use-memory=MB 选项传递给xtrabackup子进程。恢复使用内存大小

–parallel=NUMBER-OF-THREADS 选项传递给xtrabackup子进程,指定数据传输线程总数。默认为1

–throttle=IOS 选项传递给xtrabackup子进程,限制IO线程数量

–sleep=MS 选项传递给xtrabackup子进程。每拷贝1MB数据暂停多少MS时间

–compress[=LEVEL] 选项传递给xtrabackup子进程。压缩级别在0-9.1快速压缩,9最佳压缩,0不压缩。默认为1.

–include=REGEXP选项传递给xtrabackup子进程。使用正则进行匹配

–databases=LIST 指定备份数据库

–tables-file=FILE

–uncompress选项传递给xtrabackup子进程。对压缩过的InnoDB数据文件不进行压缩

–export 仅使用于prepare选项。选项传递给xtrabackup子进程。

–user=NAME

–password=WORD

–host=HOST

–port=PORT

–slave-info 备份复制从服务端,主从信息记录在ibbackup_slave_info文件中

–socket=SOCKET

–no-timestamp 不在备份根目录下创建以当前时间戳为名称的新的备份目录

–ibbackup=IBBACKUP-BINARY ibbackup二进制路径

–no-lock 禁止表级锁。全部是InnoDB引擎表和不关系二进制日志位置下使用

–scpopt=SCP-OPTIONS 指定scp参数

 

4.innobackupex备份

# ./innobackupex-1.5.1  /www/mybak/

/www/mybak/目录下将创建以时间戳为名称的目录

–stream=tar 将使用到tar4ibd进行压缩。注意:在解压时候必须加上-i参数进行解压。

 

5.恢复

# ./mysqladmin shutdown

# rm -rf innodb

# ./innobackupex-1.5.1 –apply-log /www/mybak/2011-04-11_15-28-02/

# ./innobackupex-1.5.1 –copy-back /www/mybak/2011-04-11_15-28-02/

# chown –R mysql.mysql /www/mydata

 

6.xtrabackup备份

# ./xtrabackup –defaults-file=/etc/my.cnf –backup –target-dir=/mydata/mybak/

# ls

cacti  ibdata1  nagios  test  xtrabackup_checkpoints  xtrabackup_logfile

 

7.恢复

# ./xtrabackup –defaults-file=/etc/my.cnf –prepare –target-dir=/mydata/mybak/

./xtrabackup  Ver 1.5 Rev undefined for 5.1.53 unknown-linux-gnu (x86_64)

xtrabackup: cd to /mydata/mybak/

xtrabackup: This target seems to be not prepared yet.

xtrabackup: xtrabackup_logfile detected: size=2097152, start_lsn=(732446480)

xtrabackup: Temporary instance for recovery is set as followings.

xtrabackup:   innodb_data_home_dir = ./

xtrabackup:   innodb_data_file_path = ibdata1:10M:autoextend

xtrabackup:   innodb_log_group_home_dir = ./

xtrabackup:   innodb_log_files_in_group = 1

xtrabackup:   innodb_log_file_size = 2097152

xtrabackup: Starting InnoDB instance for recovery.

xtrabackup: Using 104857600 bytes for buffer pool (set by –use-memory parameter)

InnoDB: The InnoDB memory heap is disabled

InnoDB: Mutexes and rw_locks use GCC atomic builtins

InnoDB: Compressed tables use zlib 1.2.3

InnoDB: Warning: innodb_file_io_threads is deprecated. Please use innodb_read_io_threads and innodb_write_io_threads instead

110411 16:49:25  InnoDB: highest supported file format is Barracuda.

InnoDB: Log scan progressed past the checkpoint lsn 732446480

110411 16:49:25  InnoDB: Database was not shut down normally!

InnoDB: Starting crash recovery.

InnoDB: Reading tablespace information from the .ibd files…

InnoDB: Doing recovery: scanned up to log sequence number 733504186 (56 %)

110411 16:49:25  InnoDB: Starting an apply batch of log records to the database…

InnoDB: Progress in percents: 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99

InnoDB: Apply batch completed

InnoDB: Last MySQL binlog file position 0 560768957, file name ./mysql-bin.000001

110411 16:49:25 Percona XtraDB (http://www.percona.com) 1.0.13-12.4 started; log sequence number 733504186

 

[notice (again)]

  If you use binary log and don’t use any hack of group commit,

  the binary log position seems to be:

InnoDB: Last MySQL binlog file position 0 560768957, file name ./mysql-bin.000001

 

xtrabackup: starting shutdown with innodb_fast_shutdown = 1

110411 16:49:32  InnoDB: Starting shutdown…

110411 16:49:33  InnoDB: Shutdown completed; log sequence number 733553275

# ls

cacti  ibdata1  nagios  test  xtrabackup_binlog_pos_innodb  xtrabackup_checkpoints  xtrabackup_logfile

# ./xtrabackup –defaults-file=/etc/my.cnf –prepare –target-dir=/mydata/mybak/

./xtrabackup  Ver 1.5 Rev undefined for 5.1.53 unknown-linux-gnu (x86_64)

xtrabackup: cd to /mydata/mybak/

xtrabackup: This target seems to be already prepared.

xtrabackup: notice: xtrabackup_logfile was already used to ‘–prepare’.

xtrabackup: Temporary instance for recovery is set as followings.

xtrabackup:   innodb_data_home_dir = ./

xtrabackup:   innodb_data_file_path = ibdata1:10M:autoextend

xtrabackup:   innodb_log_group_home_dir = ./

xtrabackup:   innodb_log_files_in_group = 3

xtrabackup:   innodb_log_file_size = 268435456

xtrabackup: Starting InnoDB instance for recovery.

xtrabackup: Using 104857600 bytes for buffer pool (set by –use-memory parameter)

InnoDB: The InnoDB memory heap is disabled

InnoDB: Mutexes and rw_locks use GCC atomic builtins

InnoDB: Compressed tables use zlib 1.2.3

InnoDB: Warning: innodb_file_io_threads is deprecated. Please use innodb_read_io_threads and innodb_write_io_threads instead

110411 16:49:58  InnoDB: Log file ./ib_logfile0 did not exist: new to be created

InnoDB: Setting log file ./ib_logfile0 size to 256 MB

InnoDB: Database physically writes the file full: wait…

InnoDB: Progress in MB: 100 200

110411 16:50:03  InnoDB: Log file ./ib_logfile1 did not exist: new to be created

InnoDB: Setting log file ./ib_logfile1 size to 256 MB

InnoDB: Database physically writes the file full: wait…

InnoDB: Progress in MB: 100 200

110411 16:50:09  InnoDB: Log file ./ib_logfile2 did not exist: new to be created

InnoDB: Setting log file ./ib_logfile2 size to 256 MB

InnoDB: Database physically writes the file full: wait…

InnoDB: Progress in MB: 100 200

110411 16:50:14  InnoDB: highest supported file format is Barracuda.

InnoDB: The log sequence number in ibdata files does not match

InnoDB: the log sequence number in the ib_logfiles!

110411 16:50:14  InnoDB: Database was not shut down normally!

InnoDB: Starting crash recovery.

InnoDB: Reading tablespace information from the .ibd files…

InnoDB: Last MySQL binlog file position 0 560768957, file name ./mysql-bin.000001

110411 16:50:14 Percona XtraDB (http://www.percona.com) 1.0.13-12.4 started; log sequence number 733553676

[notice (again)]

  If you use binary log and don’t use any hack of group commit,

  the binary log position seems to be:

InnoDB: Last MySQL binlog file position 0 560768957, file name ./mysql-bin.000001

xtrabackup: starting shutdown with innodb_fast_shutdown = 1

110411 16:50:16  InnoDB: Starting shutdown…

110411 16:50:16  InnoDB: Shutdown completed; log sequence number 733553676

# ls

cacti    ib_logfile0  ib_logfile2  test                          xtrabackup_checkpoints

ibdata1  ib_logfile1  nagios       xtrabackup_binlog_pos_innodb  xtrabackup_logfile

 

此时InnoDB的全部数据文件已经恢复了。将这些文件拷贝到datadir目录下。

注意:xtrabackup只是备份数据文件,不会备份数据表结构。因此恢复时,必须有对应的表结构文件。

 

# cp -r cacti /mydata/data/

# cp -r ib* /mydata/data/

# cp -r nagios /mydata/data/

# chown –R mysql.mysql /mydata/data

 

# /usr/local/mysql-5.5.11/bin/mysqld –user=mysql &

 

8.增量备份

         a.全量备份

         # ./xtrabackup –backup –target-dir=/mydata/mybak/base

         xtrabackup: The latest check point (for incremental): ‘783599510

>> log scanned up to (783977887)

xtrabackup: Stopping log copying thread.

xtrabackup: Transaction log of lsn (783374191) to (783977887) was copied.

 

# ls

cacti  ibdata1  nagios  xtrabackup_checkpoints  xtrabackup_logfile

 

# cat xtrabackup_checkpoints

backup_type = full-backuped

from_lsn = 0

to_lsn = 783599510

last_lsn = 783977887

 

b.增量备份

# ./xtrabackup –backup –target-dir=/mydata/mybak/delta –incremental-basedir=/mydata/mybak/base

incremental backup from 783599510 is enabled.

……

xtrabackup: The latest check point (for incremental): ‘793114541

>> log scanned up to (795499255)

xtrabackup: Stopping log copying thread.

xtrabackup: Transaction log of lsn (793069999) to (795499255) was copied.

 

# ls /mydata/mybak/delta/

cacti  ibdata1.delta  nagios  xtrabackup_checkpoints  xtrabackup_logfile

 

# cat /mydata/mybak/delta/xtrabackup_checkpoints

backup_type = incremental

from_lsn = 783599510

to_lsn = 793114541

last_lsn = 795499255

 

在增量备份目录下,数据文件都是以.delta结尾的。增量备份只备份上一次全量备份之后被修改过的page

 

C.恢复过程:需要分别对全量和增量备份进行恢复操作

# ./xtrabackup –prepare –target-dir=/mydata/mybak/base

# ./xtrabackup –prepare –target-dir=/mydata/mybak/base –incremental-dir=/mydata/mybak/delta

上述过程,首先将全部文件备份到/mydata/mybak/base目录下,增量备份产生的文件备份到/mydata/mybak/delta目录下。在恢复过程中,首先指定全量备份的路径,然后将增量备份应用于该全量备份。

/mydata/mybak/base目录下的文件拷贝到datadir目录下,即可。同时还需要数据表结构文件。