利用keepalived实现redis高可用

环境介绍:

master: 192.168.0.100

slave: 192.168.0.101

vip(keepvlied虚拟IP): 192.168.0.200

注:这里有一个坑

  就是在同一网段之间的Keepalived的配置文件中需要把 virtual_router_id 后面的值设置不同(范围是0-255),否则,keepalived在主从切换的时间会有问题,而且在/var/log/messages日志中

会不断的写入日志,会把服务器写爆。

virtual_router_id 每一套都需要设置不同(范围是0-255)

当 Master 与 Slave 均运作正常时, Master负责服务,Slave负责Standby;
当 Master 挂掉,Slave 正常时, Slave接管服务,同时关闭主从复制功能;
当 Master 恢复正常,则从Slave同步数据,同步数据之后关闭主从复制功能,恢复Master身份,于此同时Slave等待Master同步数据完成之后,恢复Slave身份。
然后依次循环。

需要注意的是,这样做需要在Master与Slave上都开启本地化策略,否则在互相自动切换的过程中,未开启本地化的一方会将另一方的数据清空,造成数据完全丢失

在Master和Slave上安装Keepalived

yum install -y keepalived

首先,在Master上创建如下配置文件

! Configuration File for keepalivedglobal_defs {   router_id redis100}vrrp_script chk_redis {      script "/etc/keepalived/scripts/redis_check.sh 127.0.0.1 6379"      interval 2      timeout 2     fall 3}vrrp_instance redis {     state MASTER # master set to SLAVE also     interface eth0       virtual_router_id 50      priority  150            nopreempt # 非抢占模式,注:主与从必须同时设置为BACKUP     advert_int 1authentication {   #all node must same         auth_type PASS         auth_pass 1111    }     virtual_ipaddress {  192.168.0.200/24    }    track_script {          chk_redis     }      notify_master "/etc/keepalived/scripts/redis_master.sh 127.0.0.1 192.168.0.101 6379"     notify_backup "/etc/keepalived/scripts/redis_backup.sh 127.0.0.1 192.168.0.101 6379"     notify_fault /etc/keepalived/scripts/redis_fault.sh      notify_stop /etc/keepalived/scripts/redis_stop.sh }

然后,在Slave上创建如下配置文件:

! Configuration File for keepalivedglobal_defs {   router_id redis101}vrrp_script chk_redis {      script "/etc/keepalived/scripts/redis_check.sh 127.0.0.1 6379"      interval 2     timeout 2     fall 3}vrrp_instance redis {    state BACKUP       interface eth0       virtual_router_id 50      priority  100           advert_int 1authentication {   #all node must same        auth_type PASS        auth_pass 1111    }     virtual_ipaddress {     192.168.0.200/24    }    track_script {          chk_redis     }     notify_master "/etc/keepalived/scripts/redis_master.sh 127.0.0.1 192.168.0.100 6379"    notify_backup "/etc/keepalived/scripts/redis_backup.sh 127.0.0.1 192.168.0.100 6379"    notify_fault /etc/keepalived/scripts/redis_fault.sh     notify_stop /etc/keepalived/scripts/redis_stop.sh }

在Master和Slave上创建监控Redis的脚本

 mkdir /etc/keepalived/scripts vim /etc/keepalived/scripts/redis_check.sh  #!/bin/bash ALIVE=`/usr/redis/redis-cli -h $1 -p $2 PING` LOGFILE="/var/log/keepalived-redis-check.log" echo "[CHECK]" >> $LOGFILEdate >> $LOGFILEif [ $ALIVE == "PONG" ]; then :   echo "Success: redis-cli -h $1 -p $2 PING $ALIVE" >> $LOGFILE 2>&1    exit 0 else     echo "Failed:redis-cli -h $1 -p $2 PING $ALIVE " >> $LOGFILE 2>&1    exit 1 fi

编写以下负责运作的关键脚本

notify_master /etc/keepalived/scripts/redis_master.sh
notify_backup /etc/keepalived/scripts/redis_backup.sh
notify_fault /etc/keepalived/scripts/redis_fault.sh
notify_stop /etc/keepalived/scripts/redis_stop.sh

因为Keepalived在转换状态时会依照状态来呼叫:
当进入Master状态时会呼叫notify_master
当进入Backup状态时会呼叫notify_backup
当发现异常情况时进入Fault状态呼叫notify_fault
当Keepalived程序终止时则呼叫notify_stop

首先,在Redis Master上创建notity_master与notify_backup脚本:

 cat /etc/keepalived/scripts/redis_master.sh

#!/bin/bash REDISCLI="/usr/redis/redis-cli -h $1 -p $3" LOGFILE="/var/log/keepalived-redis-state.log" echo "[master]" >> $LOGFILE date >> $LOGFILE echo "Being master...." >> $LOGFILE 2>&1  echo "Run MASTER cmd ..." >> $LOGFILE 2>&1$REDISCLI SLAVEOF $2 $3 >> $LOGFILE  sleep 10 #delay 10 s wait data async cancel syncecho "Run SLAVEOF NO ONE cmd ..." >> $LOGFILE$REDISCLI SLAVEOF NO ONE >> $LOGFILE 2>&1

cat /etc/keepalived/scripts/redis_backup.sh

#!/bin/bash REDISCLI="/usr/redis/redis-cli" LOGFILE="/var/log/keepalived-redis-state.log" echo "[backup]" >> $LOGFILE date >> $LOGFILE echo "Run SLAVEOF cmd ..." >> $LOGFILE $REDISCLI SLAVEOF $2 $3 >> $LOGFILE 2>&1 # echo "Being slave...." >> $LOGFILE 2>&1 sleep 15 #delay 15 s wait data sync exchange role

接着,在Redis Slave上创建notity_master与notify_backup脚本:

cat /etc/keepalived/scripts/redis_master.sh

#!/bin/bash REDISCLI="/usr/redis/redis-cli -h $1 -p $3" LOGFILE="/var/log/keepalived-redis-state.log" echo "[master]" >> $LOGFILE date >> $LOGFILE echo "Being master...." >> $LOGFILE 2>&1 echo "Run SLAVEOF cmd ... " >> $LOGFILE $REDISCLI SLAVEOF $2 $3 >> $LOGFILE  2>&1#echo "SLAVEOF $2 cmd can't excute ... " >> $LOGFILE sleep 10 ##delay 15 s wait data sync exchange roleecho "Run SLAVEOF NO ONE cmd ..." >> $LOGFILE$REDISCLI SLAVEOF NO ONE >> $LOGFILE 2>&1

cat /etc/keepalived/scripts/redis_backup.sh

#!/bin/bash REDISCLI="/usr/redis/redis-cli" LOGFILE="/var/log/keepalived-redis-state.log" echo "[BACKUP]" >> $LOGFILE date >> $LOGFILE echo "Being slave...." >> $LOGFILE 2>&1 echo "Run SLAVEOF cmd ..." >> $LOGFILE 2>&1$REDISCLI SLAVEOF $2 $3 >> $LOGFILE  sleep 100 #delay 10 s wait data async cancel sync exit(0)

然后在Master与Slave创建如下相同的脚本:

cat /etc/keepalived/scripts/redis_fault.sh

#!/bin/bash LOGFILE=/var/log/keepalived-redis-state.log echo "[fault]" >> $LOGFILEdate >> $LOGFILE

cat /etc/keepalived/scripts/redis_stop.sh

#!/bin/bash LOGFILE=/var/log/keepalived-redis-state.log echo "[stop]" >> $LOGFILE date >> $LOGFILE

给脚本都加上可执行权限:

chmod +x /etc/keepalived/scripts/*.sh

脚本创建完成以后,我们开始按照如下流程进行测试

1.启动Master上的Redis

$ /etc/init.d/redis start

2.启动Slave上的Redis

$ /etc/init.d/redis start

3.启动Master上的Keepalived

$ /etc/init.d/keepalived start

4.启动Slave上的Keepalived

$ /etc/init.d/keepalived start

5.尝试通过VIP连接Redis:

$ redis-cli -h 10.6.1.200 INFO

连接成功,Slave也连接上来了。

role:masterslave0:192.168.0.101,6379,online

6.尝试插入一些数据:

$ redis-cli -h 10.6.1.200 SET Hello RedisOK

从VIP读取数据

$ redis-cli -h 10.6.1.200 GET Hello"Redis"

从Master读取数据

$ redis-cli -h 10.6.1.143 GET Hello"Redis"

从Slave读取数据

$ redis-cli -h 10.6.1.144 GET Hello"Redis"

下面,模拟故障产生:
将Master上的Redis停了

本方来自:http://blog.csdn.net/huwei2003/article/details/41894197