理论介绍：
我们通常说的双机热备是指两台机器都在运行，但并不是两台机器都同时在提供服务。当提供服务的一台出现故障的时候，另外一台会马上自动接管并且提供服务，而且切换的时间非常短。MySQL双主复制，即互为Master-Slave(只有一个Master提供写操作)，可以实现数据库服务器的热备，但是一个Master宕机后不能实现动态切换。使用Keepalived，可以通过虚拟IP，实现双主对外的统一接口以及自动检查、失败切换机制，从而实现MySQL数据库的高可用方案。
环境说明：
版本 version 5.6.25-log
主1库ip: 10.219.24.25
主2库ip：10.219.24.22
vip: 10.219.24.30
os 版本： centos 6.7
已安装热备软件：xtrabackup
防火墙已关
补充：
主从复制原理： http://blog.csdn.NET/zhang123456456/article/details/72972701
MySQL 5.6安装 ：http://blog.csdn.net/zhang123456456/article/details/53608554
xtrabackup 安装： http://blog.csdn.Net/zhang123456456/article/details/72836184
全程：先搭建一主一从，然后反过来搭建，就是双主复制，最后配置 Mysql+Keepalived 故障转移的高可用环境
一、 先一主一从搭建 ，再反向搭建，配置双主复制
1、 主库参数调整
-- 停止主库mysql
[root@mysql02 ~]# service mysql stop
[root@mysql02 ~]# netstat -nltp|grep mysql|grep 3606
-- 主库创建relay log目录
[root@mysql02 full]# mkdir -p /data/mysql/relaylog/
[root@mysql02 full]# chown -R mysql:mysql /data/mysql/relaylog
-- 调整 my.cnf 参数
[root@mysql02 ~]# cat /etc/my.cnf
[client]
password = Oracle
port = 3306
socket = /data/mysql/mysql.sock
[mysqld]
server-id=25
port = 3306
socket = /data/mysql/mysql.sock
character_set_server = utf8
character_set_client = utf8
collation-server=utf8_general_ci
lower_case_table_names = 1
max_connections = 1000
datadir = /data/mysql
log_bin = /data/mysql/binarylog/binlog
log_bin_index = /data/mysql/binarylog/binlog
relay-log = /data/mysql/relaylog/relay
relay-log-index = /data/mysql/relaylog/relay
relay_log_purge = on
binlog_format = mixed
innodb_data_file_path = ibdata1:12M:autoextend
auto_increment_increment = 10
auto_increment_offset = 1
[mysql]
default-character-set = utf8
说明：
a、主库必须配置的参数
server-id （主从的server-id必须不同）、log_bin、binlog_format、relay-log、relay-log-index、relay_log_purge
auto-increment-offset、auto-increment-increment
b、log-slave-updates 意思是，中继日志执行之后，这些变化是否需要计入自己的binarylog。 当你的B服务器需要作为另外一个服务器的主服务器的时候需要打开。 就是双主互相备份，或者多主循环备份。 我们这里需要， 所以打开。
c、auto-increment-offset、auto-increment-increment 两个参数用于在 双主（多主循环）互相备份。 因为每台数据库服务器都可能在同一个表中插入数据，如果表有一个自动增长的主键，那么就会在多服务器上出现主键冲突。 解决这个问题的办法就是让每个数据库的自增主键不连续。 上图说是， 我假设需要将来可能需要10台服务器做备份， 所以auto-increment-increment 设为10. 而 auto-increment-offset=1 表示这台服务器的序号。 从1开始， 不超过auto-increment-increment。这样做之后， 我在这台服务器上插入的第一个id就是 1， 第二行的id就是 11了， 而不是2.（同理，在第二台服务器上插入的第一个id就是2， 第二行就是12， 这个后面再介绍） 这样就不会出现主键冲突了。 后面我们会演示这个id的效果。
-- 启动主库
[root@mysql02 ~]# mysqld_safe --defaults-file=/etc/my.cnf &
2、 从库参数调整
-- 停止从库mysql
[root@mysql01 ~]# service mysql stop
[root@mysql01 ~]# netstat -nltp|grep mysql|grep 3606
-- 调整 my.cnf 参数
[root@mysql01 ~]# cat /etc/my.cnf
[client]
password = oracle
port = 3306
socket = /data/mysql/mysql.sock
[mysqld]
server-id=22
port = 3306
socket = /data/mysql/mysql.sock
character_set_server = utf8
character_set_client = utf8
collation-server=utf8_general_ci
lower_case_table_names = 1
max_connections = 1000
datadir = /data/mysql
log_bin = /data/mysql/binarylog/binlog
log_bin_index = /data/mysql/binarylog/binlog
relay-log = /data/mysql/relaylog/relay
relay-log-index = /data/mysql/relaylog/relay
relay_log_purge = on
auto_increment_increment = 10
auto_increment_offset = 2
[mysql]
default-character-set = utf8
说明：从库必须配置的参数
server-id、log_bin、relay-log、relay_log_purge、auto-increment-offset、auto-increment-increment、
3、 主库备份
-- 主库备份目录
[root@mysql02 full]# pwd
/xtrabackup/full
-- 主库 innobackupex 备份
[root@mysql02 ~]# innobackupex --user=root --password=oracle --port=3606 /xtrabackup/full/
170610 17:50:23 Backup created in directory '/xtrabackup/full/2017-06-10_17-50-19/'
MySQL binlog position: filename 'binlog.000010', position '120'
....
170610 17:50:23 completed OK!
-- 查看备份 binlog编号 与 截止 position
[root@mysql02 2017-06-10_17-50-19]# cat xtrabackup_binlog_info
binlog.000010 120
4、 从库创建与主库相同的备份目录
[root@mysql01 ~]# mkdir -p /xtrabackup/full
[root@mysql01 ~]# chown -R mysql:mysql /xtrabackup/full/
5、 主库将备份 scp 到从库
[root@mysql02 full]# pwd
/xtrabackup/full
[root@mysql02 full]# scp -r 2017-06-10_17-50-19 10.219.24.22:/xtrabackup/full
6、 从库查看scp过来的备份
[root@mysql01 ~]# cd /xtrabackup/full/2017-06-10_17-50-19/
[root@mysql01 2017-06-10_17-50-19]# ll
total 12320
-rw-r-----. 1 root root 419 Jun 10 18:01 backup-my.cnf
-rw-r-----. 1 root root 12582912 Jun 10 18:01 ibdata1
drwxr-x---. 2 root root 4096 Jun 10 18:01 mysql
drwxr-x---. 2 root root 4096 Jun 10 18:01 performance_schema
drwxr-x---. 2 root root 4096 Jun 10 18:01 test
-rw-r-----. 1 root root 18 Jun 10 18:01 xtrabackup_binlog_info
-rw-r-----. 1 root root 113 Jun 10 18:01 xtrabackup_checkpoints
-rw-r-----. 1 root root 482 Jun 10 18:01 xtrabackup_info
-rw-r-----. 1 root root 2560 Jun 10 18:01 xtrabackup_logfile
7、 主库创建同步用户
mysql> GRANT replication slave ON *.* TO 'slave25'@'%' IDENTIFIED BY 'oracle';
Query OK, 0 rows affected (0.05 sec)
8、 从库恢复主库数据
-- 从库将原有datadir文件夹重命名到新位置，并创建原文件夹
[root@mysql01 ~]# mv /data/mysql /data/mysqlbak
[root@mysql01 ~]# mkdir -p /data/mysql
-- innobackupex apply-log
[root@mysql01 ~]# innobackupex --apply-log --user=oracle
--password=oracle --port=3606 /xtrabackup/full/2017-06-10_17-50-19/
-- innobackupex copy 恢复的文件到原来的数据位置
[root@mysql01 mysql]# innobackupex --defaults-file=/etc/my.cnf --user=root
--copy-back /xtrabackup/full/2017-06-10_17-50-19/
170610 18:25:11 completed OK!
-- 创建binlog目录与 relaylog 目录并赋权
[root@mysql01 ~]# mkdir -p /data/mysql/binarylog
[root@mysql01 ~]# mkdir -p /data/mysql/relaylog/
[root@mysql01 mysql]# chown -R mysql:mysql /data/mysql
9、 从库配置与检测
-- 从库启动
[root@mysql01 mysql]# mysqld_safe --defaults-file=/etc/my.cnf &
-- 从库指定与主库同步的基本信息
mysql>
change master to
master_host='10.219.24.25',
master_port=3306,
master_user='slave25',
master_password='oracle',
master_log_file='binlog.000010',
master_log_pos=120;
Query OK, 0 rows affected, 2 warnings (0.04 sec)
参数解释：
MASTER_HOST : 设置要连接的主服务器的ip地址
MASTER_USER : 设置要连接的主服务器的用户名
MASTER_PASSWORD : 设置要连接的主服务器的密码
MASTER_LOG_FILE : 设置要连接的主服务器的bin日志的日志名称
MASTER_LOG_POS : 设置要连接的主服务器的bin日志的记录位置
-- 启动slave 状态(开始监听msater的变化)
mysql> start slave;
Query OK, 0 rows affected, 1 warning (0.00 sec)
-- 查看slave的状态.
mysql> show slave status G;
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.219.24.25 #主库 IP
Master_User: slave25 # 主库复制的用户
Master_Port: 3306 # 主库 mysqld
Connect_Retry: 60
Master_Log_File: binlog.000010 #io_thread 读取主库 master_log_file
Read_Master_Log_Pos: 717 # io_thread 读取主库 master_log_pos
Relay_Log_File: relay.000002
Relay_Log_Pos: 877
Relay_Master_Log_File: binlog.000010 #sql_thread 执行主库的 master_log_file
Slave_IO_Running: Yes # 关键的，io_thread 是否 running
Slave_SQL_Running: Yes # 关键的，sql_thread 是否 running
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: 717 #sql_thread 执行主库的 master_log_pos
Relay_Log_Space: 1040
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 # 从库 的延迟
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: 25
Master_UUID: 29d68531-4cf9-11e7-8e1f-000c297c4100
Master_Info_File: /data/mysql/master.info
SQL_Delay: 0
SQL_Remaining_Delay: NULL
Slave_SQL_Running_State: Slave has read all relay log; waiting for the slave I/O thread to update it
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
1 row in set (0.00 sec)
ERROR:
No query specified
10、 主从同步检查
-- 主库
mysql> create database repl;
Query OK, 1 row affected (0.00 sec)
mysql> use repl
Database changed
mysql> create table repl (id int);
Query OK, 0 rows affected (0.02 sec)
mysql> insert into repl values(1);
Query OK, 1 row affected (0.00 sec)
-- 从库
mysql> show databases;
+--------------------+
| Database |
+--------------------+
| information_schema |
| binarylog |
| mysql |
| performance_schema |
| relaylog |
| repl |
| test |
+--------------------+
7 rows in set (0.00 sec)
mysql> use repl
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 repl;
+------+
| id |
+------+
| 1 |
+------+
1 row in set (0.00 sec) >一主一从同步成功！
##################################################
#至此A到B的复制已经配置完成，下面配置从B到A的复制。#
##################################################
声明> 下面操作中 新主库即为原从库（10.219.24.22） 新从库为原主库（10.219.24.25）
11、 新主库创建同步用户
mysql> GRANT replication slave ON *.* TO 'slave22'@'%' IDENTIFIED BY 'oracle';
Query OK, 0 rows affected (0.00 sec)
12、 新主库查看 binlog 文件号与 position 点
mysql> show master status;
+---------------+----------+--------------+------------------+-------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB | Executed_Gtid_Set |
+---------------+----------+--------------+------------------+-------------------+
| binlog.000004 | 313 | | | |
+---------------+----------+--------------+------------------+-------------------+
1 row in set (0.00 sec)
13、 新从库指定与主库同步的基本信息
mysql>
change master to
master_host='10.219.24.22',
master_port=3306,
master_user='slave22',
master_password='oracle',
master_log_file='binlog.000004',
master_log_pos=313;
Query OK, 0 rows affected, 2 warnings (0.04 sec)
14、新从库打开 slave 复制功能
mysql> start slave;
Query OK, 0 rows affected (0.00 sec)
15、 新从库检测同步复制状态
mysql> show slave status G;
*************************** 1. row ***************************
Slave_IO_State: Waiting for master to send event
Master_Host: 10.219.24.22
Master_User: slave22
Master_Port: 3306
Connect_Retry: 60
Master_Log_File: binlog.000004
Read_Master_Log_Pos: 313
Relay_Log_File: relay.000002
Relay_Log_Pos: 280
Relay_Master_Log_File: binlog.000004
Slave_IO_Running: Yes
Slave_SQL_Running: Yes
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: 313
Relay_Log_Space: 443
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: 22
Master_UUID: 70023652-4dc7-11e7-9360-000c2944297a
Master_Info_File: /data/mysql/master.info
SQL_Delay: 0
SQL_Remaining_Delay: NULL
Slave_SQL_Running_State: Slave has read all relay log; waiting for the slave I/O thread to update it
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
1 row in set (0.00 sec)
ERROR:
No query specified
-- 新从库测试数据同步状态
mysql> create database mm_repl;
Query OK, 1 row affected (0.00 sec)
mysql> use mm_repl;
Database changed
mysql> create table mm_repl(id int auto_increment,name varchar(10), primary key(id));
Query OK, 0 rows affected (0.01 sec)
mysql> insert into mm_repl(name) values("andy"),("taoYe");
Query OK, 1 row affected (0.00 sec)
mysql> select * from mm_repl;
+----+-------+
| id | name |
+----+-------+
| 1 | andy |
| 11 | taoYe |
+----+-------+
2 rows in set (0.00 sec)
-- 新主库测试数据同步状态
mysql> select * from mm_repl;
+----+-------+
| id | name |
+----+-------+
| 1 | andy |
| 11 | taoYe |
+----+-------+
2 rows in set (0.00 sec)
mysql> insert into mm_repl(name) values("andy"),("taoYe");
Query OK, 2 rows affected (0.00 sec)
Records: 2 Duplicates: 0 Warnings: 0
mysql> select * from mm_repl;
+----+-------+
| id | name |
+----+-------+
| 1 | andy |
| 11 | taoYe |
| 12 | andy |
| 22 | taoYe |
+----+-------+
4 rows in set (0.00 sec)
-- 新从库检查同步复制
mysql> select * from mm_repl;
+----+-------+
| id | name |
+----+-------+
| 1 | andy |
| 11 | taoYe |
| 12 | andy |
| 22 | taoYe |
+----+-------+
4 rows in set (0.00 sec) >主主同步测试成功
二、 配置Mysql+Keepalived故障转移的高可用环境
安装keepalived （ master1 和 master2 两台机器上都安装）
二、1.1 安装 LVS
前提：已经提前配置好本地 Yum 源 配置过程可参考> http://blog.csdn.NET/zhang123456456/article/details/56690945
1、 检查一下系统内核
安装之前，首先检查一下系统内核是否支持LVS的IPVS模块；自Linux 2.6开始，系统内核完全内置了LVS的各个模块。
[root@mysql02 ~]# modprobe -l|grep ipvs
kernel/net/netfilter/ipvs/ip_vs.ko
kernel/net/netfilter/ipvs/ip_vs_rr.ko
kernel/net/netfilter/ipvs/ip_vs_wrr.ko
kernel/net/netfilter/ipvs/ip_vs_lc.ko
....
kernel/net/netfilter/ipvs/ip_vs_pe_sip.ko
看到如上输出信息，则表明系统内核默认支持IPVS模块。
2、安装依赖包
由于是源码包安装，所以编译时需要打上如下依赖包。
# yum -y install kernel-headers glibc-headers glibc-devel libgomp gcc libstdc++-devel gcc-c++
3、 yum安装 ipvsadm-1.26-4
[root@mysql02 ~]# yum -y install kernel-devel ipvsadm
===================================================================================
Package Arch Version Repository Size
===================================================================================
Installing:
ipvsadm x86_64 1.26-4.el6 c6-media 42 k
Transaction Summary
===================================================================================
Install 1 Package(s)
Total download size: 42 k
Installed size: 78 k
Downloading Packages:
Running rpm_check_debug
Running Transaction Test
Transaction Test Succeeded
Running Transaction
Installing : ipvsadm-1.26-4.el6.x86_64 1/1
Verifying : ipvsadm-1.26-4.el6.x86_64 1/1
Installed:
ipvsadm.x86_64 0:1.26-4.el6
Complete!
4、 做个 ln
[root@mysql02 ~]# ln -sv /usr/src/kernels/2.6.32-573.el6.x86_64/ /usr/src/linux <--请自行修改 kernels 版本号
5、 查看 ipvsadm 版本信息
[root@mysql02 ~]# ipvsadm -version
ipvsadm v1.26 2008/5/15 (compiled with popt and IPVS v1.2.1) #> 能看到帮助信息，则说明IPVS安装成功。
二、1.2 安装 keepalived-1.2.24
1、官网下载地址：
[root@mysql02 ~]# wget http://www.keepalived.org/software/keepalived-1.2.24.tar.gz
2、解压安装包：
[root@mysql02 ~]# ll keepalived-1.2.24.tar.gz
-rw-r--r--. 1 root root 601873 Jun 16 23:49 keepalived-1.2.24.tar.gz
[root@mysql02 ~]# tar -zxvf keepalived-1.2.24.tar.gz
3、进入安装目录：
[root@mysql02 ~]# cd keepalived-1.2.24
4、安装keepalived:
[root@mysql03 keepalived-1.2.24]# ./configure --prefix=/usr/local/keepalived --with-kernel-dir=/usr/src/kernels/2.6.32-573.el6.x86_64/
说明：--with-kernel-dri ：指定使用内核源码中的头文件，即include目录，该参数很重要，只有使用LVS时才用此参数。
[root@mysql03 keepalived-1.2.24]# make && make install
5、复制脚本和文件：
5.1 拷贝执行文件
[root@mysql02 keepalived-1.2.24]# cp /usr/local/keepalived/sbin/keepalived /usr/sbin/
5.2 将init.d文件拷贝到etc下,加入开机启动项
[root@mysql02 keepalived-1.2.24]# cp /usr/local/keepalived/etc/rc.d/init.d/keepalived /etc/init.d/
5.3 将keepalived文件拷贝到etc下，加入网卡配置
[root@mysql02 keepalived-1.2.24]# cp /usr/local/keepalived/etc/sysconfig/keepalived /etc/sysconfig/
5.4 创建keepalived文件夹
[root@mysql02 keepalived-1.2.24]#mkdir /etc/keepalived
5.5 将keepalived配置文件拷贝到etc下
[root@mysql02 keepalived-1.2.24]# cp /usr/local/keepalived/etc/keepalived/keepalived.conf /etc/keepalived/
5.6 复制启动脚本
[root@mysql02 keepalived-1.2.24]# cp /usr/local/keepalived/sbin/keepalived /etc/keepalived/
6、加入开机启动项
[root@mysql02 keepalived-1.2.24]# chkconfig --add keepalived #添加时必须保证/etc/init.d/keepalived存在
7、 启动
[root@mysql02 keepalived-1.2.24]# chmod 755 /etc/init.d/keepalived
[root@mysql03 keepalived-1.2.24]# service keepalived start
Starting keepalived: [ OK ]
8、 关闭
[root@mysql02 keepalived-1.2.24]# service keepalived stop
Stopping keepalived: [ OK ] > OK，成功！
_________________________________________________________________________________________________________
9、 绑定虚拟IP（VIP），只在master1机器上绑定。## 说明，这一步随便做不做，不影响正常使用，作为了解。
说明：
绑定虚拟IP（VIP），这个只在master1机器上绑定即可。如下配置好keepalived后，当master1机器出现故障时，VIP资源就会被master2机器自动接管！（这一步其实可以不用这么直接在外部使用ifconfig绑定，VIP就直接在 Keepalived 的配置文件里配置就好，使用命令 ip addr 就能看出 vip；只有LVS+Keepalived 四层负载均衡的高可用环境中， vip是要在外面单独设置的（即ifconfig eth0:0 ....的方式创建vip），通过 ifconfig 可以查看出来 vip。）
[root@mysql02 keepalived-1.2.24]# ifconfig
eth0 Link encap:Ethernet HWaddr 00:0C:29:7C:41:00
inet addr:10.219.24.25 Bcast:10.255.255.255 Mask:255.0.0.0
inet6 addr: fe80::20c:29ff:fe7c:4100/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:12156 errors:0 dropped:0 overruns:0 frame:0
TX packets:2700 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:1389340 (1.3 MiB) TX bytes:901117 (879.9 KiB)
[root@mysql02 keepalived-1.2.24]# ifconfig eth0:0 10.219.24.30 broadcast 10.255.255.255 netmask 255.0.0.0 up
[root@mysql02 keepalived-1.2.24]# route add -host 10.219.24.30 dev eth0:0
[root@mysql02 keepalived-1.2.24]# ifconfig
eth0 Link encap:Ethernet HWaddr 00:0C:29:7C:41:00
inet addr:10.219.24.25 Bcast:10.255.255.255 Mask:255.0.0.0
inet6 addr: fe80::20c:29ff:fe7c:4100/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:12397 errors:0 dropped:0 overruns:0 frame:0
TX packets:2748 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:1405326 (1.3 MiB) TX bytes:906805 (885.5 KiB)
eth0:0 Link encap:Ethernet HWaddr 00:0C:29:7C:41:00
inet addr:10.219.24.30 Bcast:10.255.255.255 Mask:255.0.0.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
补充：
删除虚拟ip命令：ip addr del 10.219.24.30 dev eth0 以及 route del -host 10.219.24.30 dev eth0:0
_____________________________________________________________________________________________________________________
9、 master1 机器上的keepalived.conf配置。（下面配置中没有使用lvs的负载均衡功能，所以不需要配置虚拟服务器virtual server）
[root@mysql02 keepalived-1.2.24]# cp /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf.bak
-- 清空默认内容
[root@mysql02 keepalived-1.2.24]# > /etc/keepalived/keepalived.conf
-- 采用如下配置
[root@mysql02 keepalived-1.2.24]# cat /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
router_id MYSQL-MASTER-HA
}
vrrp_script chk_mysql_port {
script "/data/chk_mysql.sh" # mysql 健康检查脚本
interval 2 #脚本执行间隔，每2s检测一次
}
vrrp_instance VI_2 {
state BACKUP
interface eth0 #指定虚拟ip的网卡接口
virtual_router_id 54 #路由器标识，MASTER和BACKUP必须是一致的
priority 100
advert_int 1
nopreempt
authentication {
auth_type PASS
auth_pass 1111
}
virtual_ipaddress {
10.219.24.30 # vip
}
track_script {
chk_mysql_port
}
}
10、 master2机器上的keepalived配置。
master2机器上的keepalived.conf文件只修改priority为90、nopreempt不设置、real_server设置本地IP。
[root@mysql01 keepalived-1.2.24]# cp /etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf.bak
[root@mysql01 keepalived-1.2.24]# >/etc/keepalived/keepalived.conf
[root@mysql01 keepalived-1.2.24]# vim /etc/keepalived/keepalived.conf
[root@mysql01 keepalived-1.2.24]# cat /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
router_id MYSQL-MASTER-HA
}
vrrp_script chk_mysql_port {
script "/data/chk_mysql.sh"
interval 2
}
vrrp_instance VI_2 {
state BACKUP
interface eth0
virtual_router_id 54
priority 90
advert_int 1
nopreempt
authentication {
auth_type PASS
auth_pass 1111
}
virtual_ipaddress {
10.219.24.30
}
track_script {
chk_mysql_port
}
}
11、 编写切换脚本。 （master1 与 master2 都操作）
KeepAlived做心跳检测，如果Master的MySQL服务挂了(3306端口挂了),那么它就会选择自杀。Slave的KeepAlived通过心跳检测发现这个情况，就会将VIP的请求接管。
-- 检查nc能否正常使用
[root@mysql02 keepalived-1.2.24]# nc
-bash: nc: command not found
-- 未安装nc包 ，yum安装
[root@mysql02 keepalived-1.2.24]# yum -y install nc
[root@mysql02 keepalived-1.2.24]# vim /data/chk_mysql.sh
[root@mysql02 ~]# cat /data/chk_mysql.sh
#!/bin/bash
# check mysql server status
# mysql端口
PORTS="3306"
function check_ports {
for port in $PORTS;do
nc -z 127.0.0.1 $port | grep -q succeeded
[ "${PIPESTATUS[1]}" -eq 0 ] && mark=${mark}1
done
# 如果mark值为空说明端口都不通。
# 如果mark等于1，说明有端口是通的。
echo $mark
}
ret1=$(check_ports)
# 如果mysql端口不通，会尝试重启一次mysql
if [ "$ret1" != 1 ];then
service mysql stop
mysqld_safe --defaults-file=/etc/my.cnf &
q! #无用命令，只是为了跳出上面的命令
sleep 1
ret2=$(check_ports)
# 如果还是有端口不通，表示mysql服务不正常，则停掉keepalived，使VIP发生切换
[ "$ret2" != 1 ] && /etc/init.d/keepalived stop
fi
12、 脚本赋权
[root@mysql02 keepalived-1.2.24]# chmod 755 /data/chk_mysql.sh
13、 master1 与 master2 开启 keepalived 进程
-- master1
[root@mysql02 keepalived-1.2.24]# service keepalived start
-- master2
[root@mysql01 keepalived-1.2.24]# service keepalived start
14、 测试
14.1 查看 vip 情况 ，发现 vip 只在 master1 上， tcpdump 抓包，发现只有 master1 10.219.24.25 多播包 这也符合 VRRP 工作机制
-- master1
[root@mysql02 keepalived-1.2.24]# ip a l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:0c:29:7c:41:00 brd ff:ff:ff:ff:ff:ff
inet 10.219.24.25/8 brd 10.255.255.255 scope global eth0
inet 10.219.24.30/32 scope global eth0
inet6 fe80::20c:29ff:fe7c:4100/64 scope link
valid_lft forever preferred_lft forever
[root@mysql02 keepalived-1.2.24]# tcpdump -n 'host 224.0.0.18'
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on eth0, link-type EN10MB (Ethernet), capture size 65535 bytes
00:58:35.883861 IP 10.219.24.25 > 224.0.0.18: VRRPv2, Advertisement, vrid 54, prio 100, authtype simple, intvl 1s, length 20
00:58:36.885543 IP 10.219.24.25 > 224.0.0.18: VRRPv2, Advertisement, vrid 54, prio 100, authtype simple, intvl 1s, length 20
-- master2
[root@mysql01 keepalived-1.2.24]# ip a l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:0c:29:44:29:7a brd ff:ff:ff:ff:ff:ff
inet 10.219.24.22/8 brd 10.255.255.255 scope global eth0
inet6 fe80::20c:29ff:fe44:297a/64 scope link
valid_lft forever preferred_lft forever
14.2 master1 中 keepalived 关闭 ， 发现 vip 漂到了 master2 上
-- master1
[root@mysql02 keepalived-1.2.24]# service keepalived stop
Stopping keepalived: [ OK ]
-- master2
[root@mysql01 keepalived-1.2.24]# ip a l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:0c:29:44:29:7a brd ff:ff:ff:ff:ff:ff
inet 10.219.24.22/8 brd 10.255.255.255 scope global eth0
inet 10.219.24.30/32 scope global eth0
inet6 fe80::20c:29ff:fe44:297a/64 scope link
valid_lft forever preferred_lft forever
14.3 再将之前关闭的 keepalived 开启，发现 vip 未漂回来， 正好符合主备为 非抢占式。
[root@mysql02 keepalived-1.2.24]# service keepalived start
Starting keepalived: [ OK ]
[root@mysql02 keepalived-1.2.24]# ip a l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:0c:29:7c:41:00 brd ff:ff:ff:ff:ff:ff
inet 10.219.24.25/8 brd 10.255.255.255 scope global eth0
inet6 fe80::20c:29ff:fe7c:4100/64 scope link
valid_lft forever preferred_lft forever
14.4 将 master1 上的 mysql 关闭， 发现 master1 上的 keepalived 进程也被脚本 kill 了， 顺利的完成了 vip 漂回 master2 上
-- master1
[root@mysql02 keepalived-1.2.24]# ip a l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:0c:29:7c:41:00 brd ff:ff:ff:ff:ff:ff
inet 10.219.24.25/8 brd 10.255.255.255 scope global eth0
inet6 fe80::20c:29ff:fe7c:4100/64 scope link
valid_lft forever preferred_lft forever
-- master2
[root@mysql01 keepalived-1.2.24]# ip a l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:0c:29:44:29:7a brd ff:ff:ff:ff:ff:ff
inet 10.219.24.22/8 brd 10.255.255.255 scope global eth0
inet 10.219.24.30/32 scope global eth0
inet6 fe80::20c:29ff:fe44:297a/64 scope link
valid_lft forever preferred_lft forever