• xtrabackup备份mysql以及创建自动定时任务


    xtrabackup的安装
    安装依赖关系
    wget -O /etc/yum.repos.d/epel.repo http://mirrors.aliyun.com/repo/epel-7.repo yum -y install perl perl-devel libaio libaio-devel perl-Time-HiRes perl-DBD-MySQL
    下载软件包,并安装软件
    wget https://www.percona.com/downloads/XtraBackup/Percona-XtraBackup-2.4.4/binary/redhat/6/x86_64/percona-xtrabackup-24-2.4.4-1.el6.x86_64.rpm yum -y install percona-xtrabackup-24-2.4.4-1.el6.x86_64.rpm
    创建备份帐号
    grant all on *.* to smile@'localhost' identified by '123456';
    FLUSH PRIVILEGES;
    创建备份目录
    mkdir /data/mysqlbackup -p
    备份
    先进行第一次全备
    innobackupex --defaults-file=/etc/my.cnf --user=smile --password=123456 --socket=/data/mysql/mysql.sock --no-timestamp /data/mysqlbackup/xfull
    第一次增备
    innobackupex --user=smile --password=123456 --incremental --no-timestamp --incremental-basedir=/data/mysqlbackup/xfull /data/mysqlbackup/xinc1
    第二次增备
    innobackupex --user=smile --password=123456 --incremental --no-timestamp --incremental-basedir=/data/mysqlbackup/xinc1 /data/mysqlbackup/xinc2
    参数说明:

    --incremental:开启增量备份功能

    --incremental-basedir:上一次备份的路径

    恢复准备
    恢复前准备
      恢复数据前的准备(合并xtabackup_log_file和备份的物理文件)
    将redo进行重做,已提交的写到数据文件,未提交的使用undo回滚,模拟CSR的过程
     第一步:先应用全备日志(--apply-log,在全备中apply-log时,只应用redo,不应用undo)
    innobackupex --apply-log --redo-only /data/mysqlbackup/xfull/
    第二步:合并增量到全备中xinc1到full中,并且apply-log,只应用redo,不应用undo
    innobackupex --apply-log --redo-only --incremental-dir=/data/mysqlbackup/xinc1 /data/mysqlbackup/xfull/
    第三步:合并增量到全备中xinc2到full中,redo和undo都应用
    innobackupex --apply-log --incremental-dir=/data/mysqlbackup/xinc2 /data/mysqlbackup/xfull/
    第四步:整体full执行apply-log,redo和undo都应用
    innobackupex --apply-log /data/mysqlbackup/xfull/
    使用innobackupex命令进行恢复
    innobackupex --copy-back /data/mysqlbackup/xfull/
    授权:chown -R mysql:mysql /data/mysql

     

    启动数据库
    systemctl start mysqld

     

    如果启动时出现如下两种报错:

    错误1: 

    2019-06-10T11:42:10.068237+08:00 0 [ERROR] InnoDB: Unable to open undo tablespace './/undo001'.
    2019-06-10T11:42:10.068267+08:00 0 [ERROR] InnoDB: Plugin initialization aborted with error Generic error
    2019-06-10T11:42:10.668794+08:00 0 [ERROR] Plugin 'InnoDB' init function returned error.
    2019-06-10T11:42:10.668841+08:00 0 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed.
    2019-06-10T11:42:10.668848+08:00 0 [ERROR] Failed to initialize builtin plugins.
    2019-06-10T11:42:10.668862+08:00 0 [ERROR] Aborting
    2019-06-10T11:42:10.033459+08:00 0 [Note] /usr/local/mysql/bin/mysqld (mysqld 5.7.26-log) starting as process

    错误2:

    2019-06-10T11:42:10.068237+08:00 0 [ERROR] InnoDB: Unable to open undo tablespace './/undo001'.
    2019-06-10T11:42:10.068267+08:00 0 [ERROR] InnoDB: Plugin initialization aborted with error Generic error
    2019-06-10T11:42:10.668794+08:00 0 [ERROR] Plugin 'InnoDB' init function returned error.
    2019-06-10T11:42:10.668841+08:00 0 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed.
    2019-06-10T11:42:10.668848+08:00 0 [ERROR] Failed to initialize builtin plugins.
    2019-06-10T11:42:10.668862+08:00 0 [ERROR] Aborting

     

    到数据目录下删除redo undo日志 
    rm -f ib_*
    rm -f ibdata1 ibtmp1
    如果不想这样可以在my.cnf中加入如下配置,但加入需要重新初始化:
    innodb_max_undo_log_size = 128M
    innodb_undo_log_truncate = ON
    innodb_undo_logs = 128
    innodb_undo_tablespaces = 3
    innodb_purge_rseg_truncate_frequency = 10

     


    做完如上配置后,如果你的数据库不大,可以每天进行一次全备份,备份脚本如下:

     [root@db ~]# cat /server/script/mysql_full_backup.sh

    time="$(date +"back_%Y-%m-%d-%H-%M-%S")"

    innobackupex --defaults-file=/etc/my.cnf --user=mha --host=172.x.x.4 --password=123456 --socket=/server/mysql/mysqldir/mysql.sock --no-timestamp /server/mysqlbackup/xfull_${time} &>/dev/null 

    自动备份任务:

     [root@db ~]# crontab -l

    0 1 * * * sh /server/script/mysql_full_backup.sh

     每天凌晨1点整执行备份脚本

     

     

  • 相关阅读:
    阿里巴巴开源的Asynchronous I/O Design and Implementation
    maven 出现错误 -source 1.5 中不支持 diamond 运算符
    Kafka设计解析(六)- Kafka高性能架构之道
    Kafka设计解析(七)- Kafka Stream
    flink如何动态支持依赖jar包提交
    Hbase技术笔记
    windows环境:idea或者eclipse指定用户名操作hadoop集群
    HBase源码实战:BufferedMutator
    HBase工具:如何查看HBase的HFile
    HBase源码实战:CreateRandomStoreFile
  • 原文地址:https://www.cnblogs.com/--smile/p/11013555.html
Copyright © 2020-2023  润新知