码迷，mamicode.com
首页 > 其他好文 > 详细

redis哨兵加上vip

时间：2020-09-17 20:32:09      阅读：30      评论：0      收藏：0      [点我收藏+]

标签：clu   local   tor   cond   soft   ast   wget   ping   etc   

redis主从模式加上vip，业务直接连接vip，这样在主发生down时，业务不用做任何调整，但是仍然需要人工介入将从库变成主库
cluster模式，我们的php支持也不太好，并且cluster模式在某个节点down后，如果没有从库来补上，那这个slot的数据就访问不了
哨兵模式，当主节点down时，由Redis Sentinel自动完成故障发现和转移，不需要人工介入，但是业务连接的ip发生了变化，现在加上vip，判断新的主，然后vip漂到新主上，业务通过vip连接，这样业务也不用调整

环境准备：
三个节点做成一主两从，每个节点部署一个哨兵
ip为

10.99.35.209
10.99.35.215
10.99.35.216

一、安装redis(三个节点都要执行)

cd /apps
wget http://download.redis.io/releases/redis-4.0.3.tar.gz
tar xf /apps/redis-4.0.3.tar.gz -C /apps/
cd /apps/redis-4.0.3
make && make PREFIX=/usr/local/redis4.0.3 install
ln -s /usr/local/redis4.0.3/ /usr/local/redis
mkdir -p /data/redis_data/6379
cd /usr/local/redis
mkdir etc

10.99.35.209为主，10.99.35.215、10.99.35.216做成10.99.35.209的从
10.99.35.209上的配置文件如下：

bind 0.0.0.0
protected-mode no
port 6379
tcp-backlog 511
timeout 0
tcp-keepalive 300
daemonize yes
supervised no
pidfile "/data/redis_data/6379/redis_6379.pid"
loglevel notice
logfile "/data/redis_data/6379/redis_6379.log"
databases 16
always-show-logo yes
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_data/6379"
slave-serve-stale-data yes
slave-read-only yes
repl-diskless-sync no
repl-diskless-sync-delay 5
repl-disable-tcp-nodelay no
slave-priority 100
maxmemory 2gb
maxmemory-policy volatile-ttl
maxmemory-samples 3
masterauth "redis123456"
requirepass "redis123456"
lazyfree-lazy-eviction no
lazyfree-lazy-expire no
lazyfree-lazy-server-del no
slave-lazy-flush no
appendonly yes
appendfilename "appendonly.aof"
appendfsync everysec
no-appendfsync-on-rewrite no
auto-aof-rewrite-percentage 100
auto-aof-rewrite-min-size 64mb
aof-load-truncated yes
aof-use-rdb-preamble no
lua-time-limit 5000
slowlog-log-slower-than 10000
slowlog-max-len 128
latency-monitor-threshold 0
notify-keyspace-events ""
hash-max-ziplist-entries 512
hash-max-ziplist-value 64
list-max-ziplist-size -2
list-compress-depth 0
set-max-intset-entries 512
zset-max-ziplist-entries 128
zset-max-ziplist-value 64
hll-sparse-max-bytes 3000
activerehashing yes
client-output-buffer-limit normal 0 0 0
client-output-buffer-limit slave 256mb 64mb 60
client-output-buffer-limit pubsub 32mb 8mb 60
hz 10
aof-rewrite-incremental-fsync yes

10.99.35.215、10.99.35.216上的配置文件如下：

bind 0.0.0.0
protected-mode no
port 6379
tcp-backlog 511
timeout 0
tcp-keepalive 300
daemonize yes
supervised no
pidfile "/data/redis_data/6379/redis_6379.pid"
loglevel notice
logfile "/data/redis_data/6379/redis_6379.log"
databases 16
always-show-logo yes
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_data/6379"
slaveof 10.99.35.209 6379
slave-serve-stale-data yes
slave-read-only yes
repl-diskless-sync no
repl-diskless-sync-delay 5
repl-disable-tcp-nodelay no
slave-priority 100
maxmemory 2gb
maxmemory-policy volatile-ttl
maxmemory-samples 3
masterauth "redis123456"
requirepass "redis123456"
lazyfree-lazy-eviction no
lazyfree-lazy-expire no
lazyfree-lazy-server-del no
slave-lazy-flush no
appendonly yes
appendfilename "appendonly.aof"
appendfsync everysec
no-appendfsync-on-rewrite no
auto-aof-rewrite-percentage 100
auto-aof-rewrite-min-size 64mb
aof-load-truncated yes
aof-use-rdb-preamble no
lua-time-limit 5000
slowlog-log-slower-than 10000
slowlog-max-len 128
latency-monitor-threshold 0
notify-keyspace-events ""
hash-max-ziplist-entries 512
hash-max-ziplist-value 64
list-max-ziplist-size -2
list-compress-depth 0
set-max-intset-entries 512
zset-max-ziplist-entries 128
zset-max-ziplist-value 64
hll-sparse-max-bytes 3000
activerehashing yes
client-output-buffer-limit normal 0 0 0
client-output-buffer-limit slave 256mb 64mb 60
client-output-buffer-limit pubsub 32mb 8mb 60
hz 10
aof-rewrite-incremental-fsync yes

三个节点启动redis

/usr/local/redis/bin/redis-server /usr/local/redis/etc/redis_6379.conf

二、配置哨兵(三个节点都要执行)

cd /data/redis_data/
mkdir 26379
cd 26379/
vim sentinel.conf
port 26379
daemonize yes
dir /data/redis_data/26379
logfile "/data/redis_data/26379/sentinel.log"
sentinel monitor mymaster 10.99.35.209 6379 2
sentinel down-after-milliseconds mymaster 30000
sentinel parallel-syncs mymaster 1
sentinel failover-timeout mymaster 180000
sentinel auth-pass mymaster redis123456
sentinel deny-scripts-reconfig yes

启动哨兵

/usr/local/redis/bin/redis-sentinel /data/redis_data/26379/sentinel.conf

三、安装配置keepalived

cd /apps
wget http://www.keepalived.org/software/keepalived-1.3.5.tar.gz
tar -zvxf keepalived-1.3.5.tar.gz
cd keepalived-1.3.5
./configure --prefix=/usr/local/keepalived
make && make install

cd /apps/keepalived-1.3.5
cp keepalived/etc/init.d/keepalived /etc/init.d/
cp /usr/local/keepalived/etc/sysconfig/keepalived /etc/sysconfig/
cp /usr/local/keepalived/sbin/keepalived /usr/sbin/
mkdir /etc/keepalived/
cp /usr/local/keepalived/etc/keepalived/keepalived.conf /etc/keepalived/

cat /etc/keepalived/keepalived.conf
! Configuration File for keepalived
vrrp_script chk_redis_port {
    script "/etc/keepalived/chk_redis.sh" #脚本路径
    interval 2 #脚本检测频率
    weight -5 #脚本执行成功与否，权重怎么计算
    fall 2 #如果连续两次检测失败，认为节点服务不可用
    rise 1 #如果连续2次检查成功则认为节点正常
}
vrrp_script chk_redis_master {
    script "/etc/keepalived/chk_redis_master.sh"
    interval 2
    weight 10
}
vrrp_instance VI_1 {
    state MASTER
    interface eth0 #节点IP的网卡
    virtual_router_id 88 #同一个instance相同
    priority 100 # 优先级，数值越大，优先级越高
    advert_int 1
    authentication { #节点间的认证，所有的必须一致
        auth_type PASS
        auth_pass 1111
    }

    virtual_ipaddress { #VIP，自定的，我觉得和外网的IP要一个网段
        10.99.35.239
    }
    track_script { #指定前面脚本的名字
        chk_redis_port
        chk_redis_master
    }
}

四、编写简单脚本

cd /etc/keepalived/
cat chk_redis.sh
#!/bin/bash
# check redis is alived

/usr/local/redis/bin/redis-cli -h 10.99.35.209 -a redis123456 -p 6379 ping
a=`echo $?`
if [ $a -eq 1 ] ;then
    echo "/etc/init.d/keepalived stop"
    /etc/init.d/keepalived stop
fi

echo $a

cat chk_redis_master.sh
#!/bin/bash
# check redis is master

set -x

redis_master_host=`/usr/local/redis/bin/redis-cli -h 10.99.35.209 -p 26379  SENTINEL get-master-addr-by-name mymaster | awk ‘{print $1}‘ |grep ‘10.99‘`
local_host=`ip ad sh|grep "scope global eth0"|grep -v grep |awk ‘{print $2}‘|awk -F ‘/‘ ‘{print $1}‘`

if [[ $redis_master_host == $local_host ]]
then
   echo "Redis Instance is master ........"
   exit 0
else
   echo "Redis Instance is slave ........"
   exit 1
fi

五、测试
209为主，vip漂在209上，如下

[root@opsys-35-209 ~]# ip ad sh
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 mq state UP qlen 1000
    link/ether 00:0c:29:a7:dc:df brd ff:ff:ff:ff:ff:ff
    inet 10.99.35.209/24 brd 10.99.35.255 scope global eth0
    inet 10.99.35.239/32 scope global eth0
    inet6 fe80::20c:29ff:fea7:dcdf/64 scope link
       valid_lft forever preferred_lft forever

登陆哨兵查看

10.99.35.209:26379> sentinel master mymaster
1) "name"
2) "mymaster"
3) "ip"
4) "10.99.35.209"
5) "port"
6) "6379"

登陆215，查看

[root@opsys-vm5-215 26379]# ip ad sh
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 mq state UP qlen 1000
    link/ether 00:0c:29:ca:fd:8d brd ff:ff:ff:ff:ff:ff
    inet 10.99.35.215/24 brd 10.99.35.255 scope global eth0
    inet6 fe80::20c:29ff:feca:fd8d/64 scope link
       valid_lft forever preferred_lft forever

登陆216，查看

[root@opsys-vm6-216 ~]# ip ad sh
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 mq state UP qlen 1000
    link/ether 00:0c:29:06:ca:4c brd ff:ff:ff:ff:ff:ff
    inet 10.99.35.216/24 brd 10.99.35.255 scope global eth0
    inet6 fe80::20c:29ff:fe06:ca4c/64 scope link
       valid_lft forever preferred_lft forever

模拟故障，验证vip是否漂在新的master节点上
登陆209，kill redis进程

[root@opsys-35-209 ~]# ps aux|grep redis
root     10065  0.3  1.3 250408 114316 ?       Ssl  Sep08   5:06 /usr/local/redis/bin/redis-server 0.0.0.0:6379                        
root     12986  0.5  0.0 143912  7752 ?        Ssl  Sep08   7:27 /usr/local/redis/bin/redis-sentinel *:26379 [sentinel]                  
root     31970  0.0  0.0 103320   864 pts/5    S+   17:12   0:00 grep redis
[root@opsys-35-209 ~]# kill -9 10065
[root@opsys-35-209 ~]# clear
[root@opsys-35-209 ~]# ip ad sh
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 mq state UP qlen 1000
    link/ether 00:0c:29:a7:dc:df brd ff:ff:ff:ff:ff:ff
    inet 10.99.35.209/24 brd 10.99.35.255 scope global eth0
    inet6 fe80::20c:29ff:fea7:dcdf/64 scope link
       valid_lft forever preferred_lft forever

哨兵查看，发现216为主了

10.99.35.209:26379> sentinel master mymaster
1) "name"
2) "mymaster"
3) "ip"
4) "10.99.35.216"
5) "port"
6) "6379"

登陆215，查看

[root@opsys-vm5-215 26379]# ip ad sh
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 mq state UP qlen 1000
    link/ether 00:0c:29:ca:fd:8d brd ff:ff:ff:ff:ff:ff
    inet 10.99.35.215/24 brd 10.99.35.255 scope global eth0
    inet6 fe80::20c:29ff:feca:fd8d/64 scope link
       valid_lft forever preferred_lft forever

登陆216查看，发现vip漂过来了

[root@opsys-vm6-216 ~]# ip ad sh
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 mq state UP qlen 1000
    link/ether 00:0c:29:06:ca:4c brd ff:ff:ff:ff:ff:ff
    inet 10.99.35.216/24 brd 10.99.35.255 scope global eth0
    inet 10.99.35.239/32 scope global eth0
    inet6 fe80::20c:29ff:fe06:ca4c/64 scope link
       valid_lft forever preferred_lft forever

启动209，模拟节点加入后，新加入的节点是否为从，vip是否还在master节点
登陆209，启动redis，查看

/usr/local/redis/bin/redis-server /usr/local/redis/etc/redis_6379.conf
[root@opsys-35-209 ~]# ip ad sh
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 mq state UP qlen 1000
    link/ether 00:0c:29:a7:dc:df brd ff:ff:ff:ff:ff:ff
    inet 10.99.35.209/24 brd 10.99.35.255 scope global eth0
    inet6 fe80::20c:29ff:fea7:dcdf/64 scope link
       valid_lft forever preferred_lft forever

登陆215查看

[root@opsys-vm5-215 26379]# ip ad sh
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 mq state UP qlen 1000
    link/ether 00:0c:29:ca:fd:8d brd ff:ff:ff:ff:ff:ff
    inet 10.99.35.215/24 brd 10.99.35.255 scope global eth0
    inet6 fe80::20c:29ff:feca:fd8d/64 scope link
       valid_lft forever preferred_lft forever

登陆216，查看

# Replication
role:master
connected_slaves:2
slave0:ip=10.99.35.215,port=6379,state=online,offset=17460952,lag=0
slave1:ip=10.99.35.209,port=6379,state=online,offset=17460813,lag=1

[root@opsys-vm6-216 26379]# ip ad sh
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 mq state UP qlen 1000
    link/ether 00:0c:29:06:ca:4c brd ff:ff:ff:ff:ff:ff
    inet 10.99.35.216/24 brd 10.99.35.255 scope global eth0
    inet 10.99.35.239/32 scope global eth0
    inet6 fe80::20c:29ff:fe06:ca4c/64 scope link
       valid_lft forever preferred_lft forever

登陆哨兵查看

10.99.35.209:26379> SENTINEL get-master-addr-by-name mymaster
1) "10.99.35.216"
2) "6379"
10.99.35.209:26379> sentinel master mymaster
1) "name"
2) "mymaster"
3) "ip"
4) "10.99.35.216"
5) "port"
6) "6379"

10.99.35.209:26379> sentinel slaves mymaster
1)  1) "name"
    2) "10.99.35.209:6379"
    3) "ip"
    4) "10.99.35.209"
    5) "port"
    6) "6379"

2)  1) "name"
    2) "10.99.35.215:6379"
    3) "ip"
    4) "10.99.35.215"
    5) "port"
    6) "6379"

redis哨兵加上vip

标签：clu   local   tor   cond   soft   ast   wget   ping   etc   

原文地址：https://blog.51cto.com/qhd2004/2530948
踩
(0)
赞
(0)
   
举报
评论 一句话评论（0）
登录后才能评论！
分享档案
更多>
2021年07月29日 (22)
2021年07月28日 (40)
2021年07月27日 (32)
2021年07月26日 (79)
2021年07月23日 (29)
2021年07月22日 (30)
2021年07月21日 (42)
2021年07月20日 (16)
2021年07月19日 (90)
2021年07月16日 (35)
周排行
mamicode.com排行更多图片
更多
友情链接
兰亭集智   国之画   百度统计   站长统计   阿里云   chrome插件   新版天听网
关于我们 - 联系我们 - 留言反馈
© 2014 mamicode.com 版权所有  联系我们:gaon5@hotmail.com
迷上了代码！