Redis集群本身已经做到了高可用和横向扩展，但是，实际情况一些小的业务没必要上集群，单个实例就可以满足业务需求，那么我们就要想办法如何保证单个实例的高可用。

keepalived是主备模式，意味着总有一台浪费着。

下面我介绍一下redis主从+sentinel(哨兵)漂移VIP的方案

一：实验环境

#redis

10.192.203.102:6400 主库

10.192.203.107:6400 从库

10.192.203.206 VIP

#sentinel

10.192.203.102:26400 sentinel 本地节点

10.192.203.107:26400 sentinel 本地节点

10.192.203.201:26400 sentinel 仲裁节点

二：实验步骤

1：安装redis

分别在10.192.203.102，10.192.203.107，10.192.203.201上安装。

#下载源码，解压缩后编译源码。

wget http://download.redis.io/releases/redis-2.8.3.tar.gz
tar xzf redis-2.8.3.tar.gz
cd redis-2.8.3
make

#编译完成后，将redis.conf和Src目录下的redis-server、redis-benchmark、redis-cli文件拷贝到一个目录下。

mkdir /usr/local/redis
mkdir -p /data/redis/redis_sentinels
mkdir /data/redis/6400
cp redis.conf /usr/local/redis
cd src
cp redis-server /usr/local/redis
cp redis-benchmark /usr/local/redis
cp redis-cli /usr/local/redis
cd /usr/local/redis

#撰写redis配置文件

(10.192.203.102，10.192.203.107)

vi /etc/redis_6400.conf

daemonize yes
pidfile "/var/run/redis_6400.pid"
port 6400
bind 0.0.0.0
timeout 0
tcp-keepalive 60
loglevel notice
logfile "/data/redis/redis_6400.log"
maxmemory 8gb
maxmemory-policy allkeys-lru
databases 16
save 900 1
save 300 10
save 60 10000
stop-writes-on-bgsave-error yes
rdbcompression yes
rdbchecksum yes
dbfilename "dump.rdb"
dir "/data/redis/6400"
slave-serve-stale-data yes
slave-read-only yes
repl-disable-tcp-nodelay no
slave-priority 100
appendonly no
appendfilename "appendonly.aof"
appendfsync everysec
no-appendfsync-on-rewrite no
auto-aof-rewrite-percentage 100
auto-aof-rewrite-min-size 64mb
lua-time-limit 5000
slowlog-log-slower-than 10000
slowlog-max-len 128
notify-keyspace-events ""
hash-max-ziplist-entries 512
hash-max-ziplist-value 64
list-max-ziplist-entries 512
list-max-ziplist-value 64
set-max-intset-entries 512
zset-max-ziplist-entries 128

2：撰写sentinel配置文件

（在10.192.203.102，10.192.203.107，10.192.203.201上）

vi /etc/redis-sentinel6400.conf

daemonize yes
port 26400
dir "/data/redis/redis_sentinels"
pidfile "/var/run/sentinel6400.pid"
logfile "/data/redis/redis_sentinels/sentinel6400.log"
sentinel monitor master6400 10.192.203.102 6400 2
sentinel down-after-milliseconds master6400 30000
sentinel failover-timeout master6400 180000
sentinel client-reconfig-script master6400 /opt/notify_master6400.sh

3：撰写漂移VIP的脚本

（10.192.203.102，10.192.203.107）

vi /opt/notify_master6400.sh

#!/bin/bash
MASTER_IP=$6
LOCAL_IP='10.192.203.102' #从库修改为10.192.203.107
VIP='10.192.203.206'
NETMASK='24'
INTERFACE='eth0'
if [ ${MASTER_IP} = ${LOCAL_IP} ]; then
/sbin/ip addr add ${VIP}/${NETMASK} dev ${INTERFACE}
/sbin/arping -q -c 3 -A ${VIP} -I ${INTERFACE}
exit 0
else
/sbin/ip addr del ${VIP}/${NETMASK} dev ${INTERFACE}
exit 0
fi
exit 1

chmod +x /opt/notify_master6400.sh

这里大概说一下这个脚本的工作原理，sentinel在做failover的过程中会传出6个参数，分别是、、、、、、，其中第6个参数from-ip也就是新的master的ip，对应脚本中的MASTER_IP，下面的if判断大家应该都很了然了，如果MASTER_IP=LOCAL_IP，那就绑定VIP，反之删除VIP。

4：启动redis服务

（10.192.203.102，10.192.203.107）

/usr/local/redis/redis-server /etc/redis_6400.conf

5：开放防火墙端口

如果防火墙是开启状态的话，需要开放相应的端口

在三台服务器上开放端口6400,26400

vi /etc/sysconfig/iptables

添加：

-A INPUT -m state —state NEW -m tcp -p tcp—dport 6400 -j ACCEPT

-A INPUT -m state —state NEW -m tcp -p tcp—dport 26400 -j ACCEPT

#重启防火墙

service iptables restart

6：初始化主从

（10.192.203.107）

/usr/local/redis/redis-cli -p 6400 slaveof 10.192.203.102 6400

7：绑定VIP到主库

（10.192.203.102）

/sbin/ip addr add 10.192.203.206/24 dev eth0

可以使用ip addr命令进行验证，如：

[root@PC redis]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu16436 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 08:00:27:04:05:16 brd ffffff:ff
inet 10.192.203.102/24 brd 10.192.203.255 scope global eth0
inet 10.192.203.206/24 scope global secondary eth0
inet6 fe80:27ff516/64 scope link tentative dadfailed
valid_lft forever preferred_lft forever

8：启动sentinel服务

(10.192.203.102，10.192.203.107，10.192.203.201)

/usr/local/redis/redis-server /etc/redis-sentinel6400.conf —sentinel

至此，整个高可用方案已经搭建完成。

9：检查replication与Sentinel是否配置成功

#检查复制信息,确保有slave连接。

/usr/local/redis/redis-cli -h 10.192.203.102 -p 6400 info Replication
# Replication
role:master
connected_slaves:1
slave0:ip=10.192.203.107,port=6400,state=online,offset=127,lag=0
master_repl_offset:127
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:2
repl_backlog_histlen:126

#检查Sentinel信息，确保status=ok

[root@PC redis]# /usr/local/redis/redis-cli-h 10.192.203.102 -p 26400 info Sentinel
# Sentinel
sentinel_masters:1
sentinel_tilt:0
sentinel_running_scripts:0
sentinel_scripts_queue_length:0
master0:name=master6400,status=ok,address=10.192.203.102:6400,slaves=1,sentinels=3

10：验证是否实现了高可用

1：停止主库

/usr/local/redis/redis-cli -h 10.192.203.102 -p 6400 shutdown

2：看从库是否提升为主库

[root@PC ~]# /usr/local/redis/redis-cli-h 10.192.203.107 -p 6400 info Replication

# Replication
role:master
connected_slaves:0
master_repl_offset:0
repl_backlog_active:0
repl_backlog_size:1048576
repl_backlog_first_byte_offset:0
repl_backlog_histlen:0

3：验证VIP是否漂移

[root@PC ~]# ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu16436 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 qlen1000
link/ether 08:00:27:04:05:16 brd ffffff:ff
inet 10.192.203.107/24 brd 10.192.203.255 scope global eth0
inet 10.192.203.206/24 scope global secondary eth0
inet6 fe80:27ff516/64 scope link tentative dadfailed
valid_lft forever preferred_lft forever

4：查看Sentinel监控状态

[root@PC redis]# /usr/local/redis/redis-cli-h 10.192.203.102 -p 26400 info Sentinel
# Sentinel
sentinel_masters:1
sentinel_tilt:0
sentinel_running_scripts:0
sentinel_scripts_queue_length:0
master0:name=master6400,status=ok,address=10.192.203.107:6400,slaves=1,sentinels=3

说明漂移成功。