Redis主从，读写分离、HA配置

Redis的安装与配置：[url]http://donald-draper.iteye.com/blog/2346163[/url]
Redis配置文件详解：[url]http://donald-draper.iteye.com/blog/2346275[/url]
Reid高可用Sentinel配置文件详解：[url]http://donald-draper.iteye.com/blog/2346606[/url]
Redis 的 Sentinel 文档：[url]http://www.redis.cn/topics/sentinel.html[/url]
redis的主从复制，读写分离，主从切换:[url]http://www.tuicool.com/articles/3MneIbz[/url]
Redis主从同步，读写分离:[url]http://www.tuicool.com/articles/vyiUveq[/url]

环境Centos7，redis-3.0.5
Master-Centos7
ip 运行进程 内存 硬盘
192.168.126.128 redis sentinel 2G 30G
Slave-Centos7
192.168.126.133 redis sentinel 1.5G 30G
192.168.126.138 redis sentinel 1.5G 30G
Master的配置，我们在前文中已讲，这里不再说了，直接配置slave
#拷贝安装文件到slave

[redis@agent133 Downloads]$ scp 192.168.126.128：/home/redis/Downloads/redis-3.0.5.tar.gz
[redis@agent133 Downloads]$ ls
redis-3.0.5.tar.gz

解压，按照我们前面的redis安装配置文章，安装redis
###从Mater拷贝redis配置文件
[code="java"]# cd /usr/local/redis/conf/
# ls -al
total 0
drwxr-xr-x. 2 root root 6 Dec 21 17:12 .
drwxr-xr-x. 4 root root 27 Dec 21 17:12 ..
# scp 192.168.126.128:/usr/local/redis/conf/redis.conf .
The authenticity of host '192.168.126.128 (192.168.126.128)' can't be established.
ECDSA key fingerprint is 75:80:da:08:ea:2d:9d:3d:da:7c:b1:1f:71:7d:b1:a5.
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added '192.168.126.128' (ECDSA) to the list of known hosts.
root@192.168.126.128's password:
redis.conf 100% 41KB 40.6KB/s 00:00
# ls -al
total 44
drwxr-xr-x. 2 root root 23 Dec 21 17:13 .
drwxr-xr-x. 4 root root 27 Dec 21 17:12 ..
-rw-r--r--. 1 root root 41584 Dec 21 17:13 redis.conf[/code]
拷贝数据目录及日志目录/redis
[code="java"]
# cd /
# ls
bin boot dev etc home lib lib64 media mnt mysql opt proc root run sbin srv sys tmp usr var zabbix
# scp -r 192.168.126.128:/redis .
root@192.168.126.128's password:
redis.log 100% 19KB 19.1KB/s 00:00
redis.sh 100% 1338 1.3KB/s 00:00
appendonly.aof 100% 245 0.2KB/s 00:00
dump.rdb 100% 18 0.0KB/s 00:00
# ls -al
total 48
dr-xr-xr-x. 20 root root 4096 Dec 21 17:17 .
dr-xr-xr-x. 20 root root 4096 Dec 21 17:17 ..
lrwxrwxrwx. 1 root root 7 Jun 24 17:17 bin -> usr/bin
dr-xr-xr-x. 4 root root 4096 Jun 27 10:54 boot
drwxr-xr-x. 20 root root 3280 Dec 21 16:53 dev
drwxr-xr-x. 142 root root 12288 Dec 21 16:59 etc
drwxr-xr-x. 6 root root 57 Dec 21 10:27 home
lrwxrwxrwx. 1 root root 7 Jun 24 17:17 lib -> usr/lib
lrwxrwxrwx. 1 root root 9 Jun 24 17:17 lib64 -> usr/lib64
drwxr-xr-x. 2 root root 6 Aug 12 2015 media
drwxr-xr-x. 2 root root 6 Aug 12 2015 mnt
drwxr-xr-x. 2 mysql root 6 Jun 29 17:36 mysql
drwxr-xr-x. 3 root root 15 Aug 12 2015 opt
dr-xr-xr-x. 450 root root 0 Dec 21 16:52 proc
drwxr-xr-x. 5 root root 38 Dec 21 17:17 redis
dr-xr-x---. 5 root root 4096 Dec 21 17:13 root
drwxr-xr-x. 39 root root 1180 Dec 21 16:59 run
lrwxrwxrwx. 1 root root 8 Jun 24 17:17 sbin -> usr/sbin
drwxr-xr-x. 2 root root 6 Aug 12 2015 srv
dr-xr-xr-x. 13 root root 0 Dec 21 16:52 sys
drwxrwxrwt. 17 root root 4096 Dec 21 17:12 tmp
drwxr-xr-x. 13 root root 4096 Jul 4 11:41 usr
drwxr-xr-x. 23 root root 4096 Dec 21 16:52 var
drwxrwxr-x. 4 zabbix mysql 4096 Aug 11 15:58 zabbix
# [/code]

##清除原先的数据文件及日志文件
[code="java"]# cd redis/
# ls
bin data logs
# cd data/
# ls -al
total 8
drwxr-xr-x. 2 root root 42 Dec 21 17:17 .
drwxr-xr-x. 5 root root 38 Dec 21 17:17 ..
-rw-r--r--. 1 root root 245 Dec 21 17:17 appendonly.aof
-rw-r--r--. 1 root root 18 Dec 21 17:17 dump.rdb
# rm ./*
rm: remove regular file ‘./appendonly.aof’? y
rm: remove regular file ‘./dump.rdb’? y
# ls -al
total 0
drwxr-xr-x. 2 root root 6 Dec 21 17:18 .
drwxr-xr-x. 5 root root 38 Dec 21 17:17 ..
# cd ..
# ls
bin data logs
# cd logs/
# ls -al
total 20
drwxr-xr-x. 2 root root 22 Dec 21 17:17 .
drwxr-xr-x. 5 root root 38 Dec 21 17:17 ..
-rw-r--r--. 1 root root 19546 Dec 21 17:17 redis.log
# > redis.log
# tail redis.log
# [/code]


修改Slave1配置
[code="java"]# cd conf/
# vim redis.conf
#

################################# REPLICATION #################################

# Master-Slave replication. Use slaveof to make a Redis instance a copy of
# another Redis server. A few things to understand ASAP about Redis replication.
#
# 1) Redis replication is asynchronous, but you can configure a master to
# stop accepting writes if it appears to be not connected with at least
# a given number of slaves.
# 2) Redis slaves are able to perform a partial resynchronization with the
# master if the replication link is lost for a relatively small amount of
# time. You may want to configure the replication backlog size (see the next
# sections of this file) with a sensible value depending on your needs.
# 3) Replication is automatic and does not need user intervention. After a
# network partition slaves automatically try to reconnect to masters
# and resynchronize with them.
# master的ip和port
slaveof 192.168.126.128 6379

# If the master is password protected (using the "requirepass" configuration
# directive below) it is possible to tell the slave to authenticate before
# starting the replication synchronization process, otherwise the master will
# refuse the slave request.
# master验证密码
masterauth redis


# Since Redis 2.6 by default slaves are read-only.
#
# Note: read only slaves are not designed to be exposed to untrusted clients
# on the internet. It's just a protection layer against misuse of the instance.
# Still a read only slave exports by default all the administrative commands
# such as CONFIG, DEBUG, and so forth. To a limited extent you can improve
# security of read only slaves using 'rename-command' to shadow all the
# administrative / dangerous commands.
##slave只读
slave-read-only yes
# A slave with a low priority number is considered better for promotion, so
# for instance if there are three slaves with priority 10, 100, 25 Sentinel will
# pick the one with priority 10, that is the lowest.
#
# However a special priority of 0 marks the slave as not able to perform the
# role of master, so a slave with priority of 0 will never be selected by
# Redis Sentinel for promotion.
#
# By default the priority is 100.
##slave优先级，优先级越低，当master宕机时，成为master的可能性越大
slave-priority 90[/code]
[color=red]##注意slaves的优先级[/color]


修改Master文件

# It is possible for a master to stop accepting writes if there are less than
# N slaves connected, having a lag less or equal than M seconds.
#
# The N slaves need to be in "online" state.
#
# The lag in seconds, that must be <= the specified value, is calculated from
# the last ping received from the slave, that is usually sent every second.
#
# This option does not GUARANTEE that N replicas will accept the write, but
# will limit the window of exposure for lost writes in case not enough slaves
# are available, to the specified number of seconds.
#
# For example to require at least 3 slaves with a lag <= 10 seconds use:
#
###为了保证集群的一致性，master必须保证salve的在线数量和握手时间满足以下条件才
接受写请求
min-slaves-to-write 1
min-slaves-max-lag 10

为了保证集群的高可用性，我们在Master和slaves上，启动Sentinel，功能与类似与zookeeper。
#从redis安装包中，拷贝sentinel配置文件到 /usr/local/redis/conf/
[code="java"]# cp sentinel.conf /usr/local/redis/conf/
# cd /usr/local/redis/conf/
# ls -al
total 52
drwxr-xr-x 2 root root 43 Dec 21 18:20 .
drwxr-xr-x 4 root root 27 Dec 19 18:16 ..
-rw-r--r-- 1 root root 41580 Dec 21 17:40 redis.conf
-rw-r--r-- 1 root root 7109 Dec 21 18:20 sentinel.conf[/code]

#创建sentinel工作目录
[code="java"]# mkdir sentinel
# ls -al
total 4
drwxr-xr-x 6 root root 53 Dec 21 18:29 .
dr-xr-xr-x. 19 root root 4096 Dec 20 15:16 ..
drwxr-xr-x 2 root root 21 Dec 20 15:37 bin
drwxr-xr-x 2 root root 42 Dec 21 10:52 data
drwxr-xr-x 2 root root 22 Dec 20 15:19 logs
drwxr-xr-x 2 root root 6 Dec 21 18:29 sentinel

# [/code]

修改sentinel.conf
# vim sentinel.conf
具体如下，具体下面每个配置的意思，我们前文sentinel配置文件篇中已讲，这里不再赘述，其中zabbix是集群中master-128的主机名，

# port <sentinel-port>
# The port that this sentinel instance will run on
port 26379
dir /redis/sentinel
sentinel monitor zabbix 192.168.126.128 6379 2
sentinel auth-pass zabbix redis
sentinel down-after-milliseconds zabbix 30000
sentinel parallel-syncs zabbix 1
sentinel failover-timeout zabbix 180000

#由于我的操作系统是centos，有防火墙，为了Master与slave能够通信，
将6379，与26379两个端口，解除防护，如下，具体可参考相关博文：
[url]http://donald-draper.iteye.com/blog/2315696[/url]
[code="java"]
# vim /etc/sysconfig/iptables
# cat /etc/sysconfig/iptables
# sample configuration for iptables service
# you can edit this manually or use system-config-firewall
# please do not ask us to add additional ports/services to this default configuration
*filter
:INPUT ACCEPT [0:0]
:FORWARD ACCEPT [0:0]
:OUTPUT ACCEPT [0:0]
-A INPUT -m state --state RELATED,ESTABLISHED -j ACCEPT
-A INPUT -p icmp -j ACCEPT
-A INPUT -i lo -j ACCEPT
-A INPUT -p tcp -m state --state NEW -m tcp --dport 22 -j ACCEPT
-A INPUT -p tcp -m state --state NEW -m tcp --dport 80 -j ACCEPT
-A INPUT -p tcp -m state --state NEW -m tcp --dport 25 -j ACCEPT
-A INPUT -p tcp -m state --state NEW -m tcp --dport 3306 -j ACCEPT
-A INPUT -p tcp -m state --state NEW -m tcp --dport 10051 -j ACCEPT
-A INPUT -p tcp -m state --state NEW -m tcp --dport 6379 -j ACCEPT
-A INPUT -p tcp -m state --state NEW -m tcp --dport 26379 -j ACCEPT
-A INPUT -j REJECT --reject-with icmp-host-prohibited
-A FORWARD -j REJECT --reject-with icmp-host-prohibited
COMMIT[/code]
##在slaves上做同样的工作，再启动集群前，确保所有机器6379，26379端口畅通，
如果以上方法解决不了，直接iptables -F
#
至此sentinel配置完毕，将master的sentinel配置文件拷贝到Slaves上
[code="java"]# cd conf/
# ls
redis.conf
# scp 192.168.126.128:/usr/local/redis/conf/sentinel.conf .
root@192.168.126.128's password:
sentinel.conf 100% 7094 6.9KB/s 00:00
# ls -al
total 52
drwxr-xr-x. 2 root root 43 Dec 21 18:40 .
drwxr-xr-x. 4 root root 27 Dec 21 17:12 ..
-rw-r--r--. 1 root root 41460 Dec 21 17:37 redis.conf
-rw-r--r--. 1 root root 7094 Dec 21 18:40 sentinel.conf
#
[/code]


启动master redis
[code="java"]# cd bin/
# ls
redis.sh
# ./redis.sh start
start redis-server...runing
# [/code]

启动master sentinel

[code="java"]#

[1] 7365
# 7365:X 21 Dec 18:44:48.955 * Increased maximum number of open files to 10032 (it was originally set to 1024).
_._
_.-``__ ''-._
_.-`` `. `_. ''-._ Redis 3.0.5 (00000000/0) 64 bit
.-`` .-```. ```\/ _.,_ ''-._
( ' , .-` | `, ) Running in sentinel mode
|`-._`-...-` __...-.``-._|'` _.-'| Port: 26379
| `-._ `._ / _.-' | PID: 7365
`-._ `-._ `-./ _.-' _.-'
|`-._`-._ `-.__.-' _.-'_.-'|
| `-._`-._ _.-'_.-' | http://redis.io
`-._ `-._`-.__.-'_.-' _.-'
|`-._`-._ `-.__.-' _.-'_.-'|
| `-._`-._ _.-'_.-' |
`-._ `-._`-.__.-'_.-' _.-'
`-._ `-.__.-' _.-'
`-._ _.-'
`-.__.-'

7365:X 21 Dec 18:44:48.956 # WARNING: The TCP backlog setting of 511 cannot be enforced because /proc/sys/net/core/somaxconn is set to the lower value of 128.
7365:X 21 Dec 18:44:48.956 # Sentinel runid is b4354571b53ffdbc7369913f3c2c87a31f2a22a4
7365:X 21 Dec 18:44:48.956 # +monitor master zabbix 192.168.126.128 6379 quorum 2[/code]

连接Master
[code="java"]#redis-cli -h localhost -p 6379 -a redis
localhost:6379> monitor
OK
1482317315.788890 [0 192.168.126.128:40190] "PING"
1482317315.843660 [0 192.168.126.128:40190] "PUBLISH" "__sentinel__:hello" "192.168.126.128,26379,b4354571b53ffdbc7369913f3c2c87a31f2a22a4,0,zabbix,192.168.126.128,6379,0"
1482317316.800317 [0 192.168.126.128:40190] "PING"
1482317317.836785 [0 192.168.126.128:40190] "PING"
1482317317.928621 [0 192.168.126.128:40190] "PUBLISH" "__sentinel__:hello" "192.168.126.128,26379,b4354571b53ffdbc7369913f3c2c87a31f2a22a4,0,zabbix,192.168.126.128,6379,0"
1482317318.871309 [0 192.168.126.128:40190] "PING"
1482317319.950063 [0 192.168.126.128:40190] "INFO"[/code]
可以看出sentinel正在监控Master

##连接sentinel查看状态
[code="java"]# redis-cli -p 26379
127.0.0.1:26379> info
# Server
redis_version:3.0.5
redis_git_sha1:00000000
redis_git_dirty:0
redis_build_id:2b4372a303619568
redis_mode:sentinel
os:Linux 3.10.0-327.22.2.el7.x86_64 x86_64
arch_bits:64
multiplexing_api:epoll
gcc_version:4.8.5
process_id:7365
run_id:b4354571b53ffdbc7369913f3c2c87a31f2a22a4
tcp_port:26379
uptime_in_seconds:1211
uptime_in_days:0
hz:18
lru_clock:5923291
config_file:/usr/local/redis/conf/sentinel.conf

# Sentinel
sentinel_masters:1
sentinel_tilt:0
sentinel_running_scripts:0
sentinel_scripts_queue_length:0
master0:name=zabbix,status=ok,address=192.168.126.128:6379,slaves=0,sentinels=1
127.0.0.1:26379> [/code]


##查看master状态：

127.0.0.1:6379> info
# Replication
role:master
connected_slaves:1
min_slaves_good_slaves:1
slave0:ip=192.168.126.133,port=6379,state=online,offset=59413,lag=1
master_repl_offset:59556
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:2
repl_backlog_histlen:59555

再slaves启动redis和sentinel
[code="java"]
# ls
bin data logs sentinel
# cd bin/
# ls
redis.sh
# ./redis.sh start
start redis-server...runing
# netstat -ntlp | grep 6379
tcp 0 0 0.0.0.0:6379 0.0.0.0:* LISTEN 4282/redis-server *
tcp6 0 0 :::6379 :::* LISTEN 4282/redis-server *
# redis-sentinel /usr/local/redis/conf/sentinel.conf &
[1] 4288
# 4288:X 21 Dec 19:12:32.938 * Increased maximum number of open files to 10032 (it was originally set to 1024).
_._
_.-``__ ''-._
_.-`` `. `_. ''-._ Redis 3.0.5 (00000000/0) 64 bit
.-`` .-```. ```\/ _.,_ ''-._
( ' , .-` | `, ) Running in sentinel mode
|`-._`-...-` __...-.``-._|'` _.-'| Port: 26379
| `-._ `._ / _.-' | PID: 4288
`-._ `-._ `-./ _.-' _.-'
|`-._`-._ `-.__.-' _.-'_.-'|
| `-._`-._ _.-'_.-' | http://redis.io
`-._ `-._`-.__.-'_.-' _.-'
|`-._`-._ `-.__.-' _.-'_.-'|
| `-._`-._ _.-'_.-' |
`-._ `-._`-.__.-'_.-' _.-'
`-._ `-.__.-' _.-'
`-._ _.-'
`-.__.-'

4288:X 21 Dec 19:12:32.940 # WARNING: The TCP backlog setting of 511 cannot be enforced because /proc/sys/net/core/somaxconn is set to the lower value of 128.
4288:X 21 Dec 19:12:32.940 # Sentinel runid is 17607179ad13102b1561dcd6a999cff732ac0f98
4288:X 21 Dec 19:12:32.940 # +monitor master zabbix 192.168.126.128 6379 quorum 2
4288:X 21 Dec 19:12:33.944 * +slave slave 192.168.126.133:6379 192.168.126.133 6379 @ zabbix 192.168.126.128 6379
4288:X 21 Dec 19:12:35.052 * +sentinel sentinel 192.168.126.128:26379 192.168.126.128 26379 @ zabbix 192.168.126.128 6379
#
[/code]

#连接Sentinel
[code="java"]# redis-cli -p 26379
127.0.0.1:26379> info
# Server
redis_version:3.0.5
redis_git_sha1:00000000
redis_git_dirty:0
redis_build_id:e0daa82ba7583198
redis_mode:sentinel
os:Linux 3.10.0-327.22.2.el7.x86_64 x86_64
arch_bits:64
multiplexing_api:epoll
gcc_version:4.8.5
process_id:4288
run_id:17607179ad13102b1561dcd6a999cff732ac0f98
tcp_port:26379
uptime_in_seconds:77
uptime_in_days:0
hz:11
lru_clock:5923821
config_file:/usr/local/redis/conf/sentinel.conf

# Sentinel
sentinel_masters:1
sentinel_tilt:0
sentinel_running_scripts:0
sentinel_scripts_queue_length:0
master0:name=zabbix,status=ok,address=192.168.126.128:6379,slaves=1,sentinels=2
127.0.0.1:26379> [/code]

在master的monitor界面，可以看到slave上的Sentinel发来的握手信息

482318873.513009 [0 192.168.126.133:52522] "PING"
1482318873.596223 [0 192.168.126.133:52522] "PUBLISH" "__sentinel__:hello" "192.168.126.133,26379,17607179ad13102b1561dcd6a999cff732ac0f98,0,zabbix,192.168.126.128,6379,0"

监控slave的redis实例，与master的握手信息

[redis@agent133 Desktop]$ redis-cli -p 6379 -a redis
127.0.0.1:6379> monitor
OK
1482318949.276759 [0 192.168.126.133:49624] "PING"
1482318949.470154 [0 192.168.126.128:6379] "PUBLISH" "__sentinel__:hello" "192.168.126.133,26379,17607179ad13102b1561dcd6a999cff732ac0f98,0,zabbix,192.168.126.128,6379,0"
1482318949.822874 [0 192.168.126.128:42838] "PING"
1482318950.128768 [0 192.168.126.128:6379] "PUBLISH" "__sentinel__:hello" "192.168.126.128,26379,b4354571b53ffdbc7369913f3c2c87a31f2a22a4,0,zabbix,192.168.126.128,6379,0"
1482318950.300715 [0 192.168.126.133:49624] "PING"
1482318950.740239 [0 192.168.126.133:49624] "PUBLIS

#查看slave状态

[redis@agent133 Desktop]$ redis-cli -p 6379 -a redis
127.0.0.1:6379> info 
# Replication
role:slave
master_host:192.168.126.128
master_port:6379
master_link_status:up
master_last_io_seconds_ago:1
master_sync_in_progress:0
slave_repl_offset:43672
slave_priority:90
slave_read_only:1
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

在master上，添加键

localhost:6379> set name donald
OK
localhost:6379> get name 
"donald"
localhost:6379>

在slaves，我们可以看到

127.0.0.1:6379> get name
"donald"
127.0.0.1:6379> 

在master上连接sentinel，查看master状态

127.0.0.1:26379> sentinel masters
1)  1) "name"
    2) "zabbix"
    3) "ip"
    4) "192.168.126.128"
    5) "port"
    6) "6379"
    7) "runid"
    8) "239033187f3fd756c470f7f31b5e49efbf7c603d"
    9) "flags"
   10) "master"
   11) "pending-commands"
   12) "0"
   13) "last-ping-sent"
   14) "0"
   15) "last-ok-ping-reply"
   16) "264"
   17) "last-ping-reply"
   18) "264"
   19) "down-after-milliseconds"
   20) "30000"
   21) "info-refresh"
   22) "1861"
   23) "role-reported"
   24) "master"
   25) "role-reported-time"
   26) "2472617"
   27) "config-epoch"
   28) "0"
   29) "num-slaves"
   30) "2"
   31) "num-other-sentinels"
   32) "1"
   33) "quorum"
   34) "2"
   35) "failover-timeout"
   36) "180000"
   37) "parallel-syncs"
   38) "1"

#在128 Master 查看slave状态

127.0.0.1:26379> sentinel slaves zabbix
1)  1) "name"
    2) "192.168.126.138:6379"
    3) "ip"
    4) "192.168.126.138"
    5) "port"
    6) "6379"
    7) "runid"
    8) "de9d33c23dc969932098503a28d7ac2cf2e67df6"
    9) "flags"
   10) "slave"
   11) "pending-commands"
   12) "0"
   13) "last-ping-sent"
   14) "0"
   15) "last-ok-ping-reply"
   16) "21"
   17) "last-ping-reply"
   18) "21"
   19) "down-after-milliseconds"
   20) "30000"
   21) "info-refresh"
   22) "7680"
   23) "role-reported"
   24) "slave"
   25) "role-reported-time"
   26) "198553"
   27) "master-link-down-time"
   28) "0"
   29) "master-link-status"
   30) "ok"
   31) "master-host"
   32) "192.168.126.128"
   33) "master-port"
   34) "6379"
   35) "slave-priority"
   36) "100"
   37) "slave-repl-offset"
   38) "119118"
2)  1) "name"
    2) "192.168.126.133:6379"
    3) "ip"
    4) "192.168.126.133"
    5) "port"
    6) "6379"
    7) "runid"
    8) "47c588eabdc0ec1fe29d40be35b612ac82980b44"
    9) "flags"
   10) "slave"
   11) "pending-commands"
   12) "0"
   13) "last-ping-sent"
   14) "0"
   15) "last-ok-ping-reply"
   16) "21"
   17) "last-ping-reply"
   18) "21"
   19) "down-after-milliseconds"
   20) "30000"
   21) "info-refresh"
   22) "7679"
   23) "role-reported"
   24) "slave"
   25) "role-reported-time"
   26) "881369"
   27) "master-link-down-time"
   28) "0"
   29) "master-link-status"
   30) "ok"
   31) "master-host"
   32) "192.168.126.128"
   33) "master-port"
   34) "6379"
   35) "slave-priority"
   36) "90"
   37) "slave-repl-offset"
   38) "119118"
127.0.0.1:26379> 

##当我们关闭138 slave redis实例时，master的sentinel会有相应的信息

127.0.0.1:26379> 7365:X 21 Dec 19:28:33.155 # +sdown slave 192.168.126.138:6379 192.168.126.138 6379 @ zabbix 192.168.126.128 6379
7365:X 21 Dec 19:28:52.068 * +reboot slave 192.168.126.138:6379 192.168.126.138 6379 @ zabbix 192.168.126.128 6379
7365:X 21 Dec 19:28:52.134 # -sdown slave 192.168.126.138:6379 192.168.126.138 6379 @ zabbix 192.168.126.128 6379

在138上连接sentinel，查看状态：

# Server
redis_version:3.0.5
redis_git_sha1:00000000
redis_git_dirty:0
redis_build_id:e0daa82ba7583198
redis_mode:sentinel
os:Linux 3.10.0-327.22.2.el7.x86_64 x86_64
arch_bits:64
multiplexing_api:epoll
gcc_version:4.8.5
process_id:4847
run_id:6183fe493327f4eff1d0d4971fce11afecc96408
tcp_port:26379
uptime_in_seconds:160
uptime_in_days:0
hz:11
lru_clock:5925146
config_file:/usr/local/redis/conf/sentinel.conf

# Sentinel
sentinel_masters:1
sentinel_tilt:0
sentinel_running_scripts:0
sentinel_scripts_queue_length:0
master0:name=zabbix,status=ok,address=192.168.126.128:6379,slaves=2,sentinels=3
127.0.0.1:26379> 

可以看出slaves为2,sentinels为3


现在我们来测试一下HA，关闭Master

localhost:6379> shutdown
not connected> 

在128上sentinel显示选举master和更新配置信息

127.0.0.1:26379> 7365:X 21 Dec 19:33:16.433 * +sentinel sentinel 192.168.126.138:26379 192.168.126.138 26379 @ zabbix 192.168.126.128 6379
7365:X 21 Dec 19:41:12.576 # +sdown master zabbix 192.168.126.128 6379
7365:X 21 Dec 19:41:12.655 # +new-epoch 1
##选择master
7365:X 21 Dec 19:41:12.657 # +vote-for-leader 6183fe493327f4eff1d0d4971fce11afecc96408 1
7365:X 21 Dec 19:41:12.661 # +odown master zabbix 192.168.126.128 6379 #quorum 3/2
7365:X 21 Dec 19:41:12.661 # Next failover delay: I will not start a failover before Wed Dec 21 19:47:13 2016
##更新slave 138配置
7365:X 21 Dec 19:41:13.762 # +config-update-from sentinel 192.168.126.138:26379 192.168.126.138 26379 @ zabbix 192.168.126.128 6379
##将133切换为Master，128切换为slave，由于133的优先级为90,138优先级为100，所以选举133为master
7365:X 21 Dec 19:41:13.762 # +switch-master zabbix 192.168.126.128 6379 192.168.126.133 6379
7365:X 21 Dec 19:41:13.762 * +slave slave 192.168.126.138:6379 192.168.126.138 6379 @ zabbix 192.168.126.133 6379
7365:X 21 Dec 19:41:13.762 * +slave slave 192.168.126.128:6379 192.168.126.128 6379 @ zabbix 192.168.126.133 6379

在133，查看sever信息：

[redis@agent133 Desktop]$ redis-cli -p 6379 -a redis
# Replication
role:master
connected_slaves:1
slave0:ip=192.168.126.138,port=6379,state=online,offset=65952,lag=0
master_repl_offset:65952
repl_backlog_active:1
repl_backlog_size:1048576
repl_backlog_first_byte_offset:2
repl_backlog_histlen:65951


db0:keys=1,expires=0,avg_ttl=0
127.0.0.1:6379> 

128重启redis时，sentinel显示convert-to-slave

7365:X 21 Dec 19:45:55.555 * +convert-to-slave slave 192.168.126.128:6379 192.168.126.128 6379 @ zabbix 192.168.126.133 6379

在133上查看sentinel master与slaves状态

127.0.0.1:26379> 
127.0.0.1:26379> sentinel masters
1)  1) "name"
    2) "zabbix"
    3) "ip"
    4) "192.168.126.133"
    5) "port"
    6) "6379"
    7) "runid"
    8) "47c588eabdc0ec1fe29d40be35b612ac82980b44"
    9) "flags"
   10) "master"
   11) "pending-commands"
   12) "0"
   13) "last-ping-sent"
   14) "0"
   15) "last-ok-ping-reply"
   16) "288"
   17) "last-ping-reply"
   18) "288"
   19) "down-after-milliseconds"
   20) "30000"
   21) "info-refresh"
   22) "4219"
   23) "role-reported"
   24) "master"
   25) "role-reported-time"
   26) "546706"
   27) "config-epoch"
   28) "1"
   29) "num-slaves"
   30) "2"
   31) "num-other-sentinels"
   32) "2"
   33) "quorum"
   34) "2"
   35) "failover-timeout"
   36) "180000"
   37) "parallel-syncs"
   38) "1"
127.0.0.1:26379> sentinel slaves zabbix
1)  1) "name"
    2) "192.168.126.128:6379"
    3) "ip"
    4) "192.168.126.128"
    5) "port"
    6) "6379"
    7) "runid"
    8) "71c8aab6f676f7d3e177a9c6321fd91a0d47fef6"
    9) "flags"
   10) "slave"
   11) "pending-commands"
   12) "0"
   13) "last-ping-sent"
   14) "0"
   15) "last-ok-ping-reply"
   16) "610"
   17) "last-ping-reply"
   18) "610"
   19) "down-after-milliseconds"
   20) "30000"
   21) "info-refresh"
   22) "5103"
   23) "role-reported"
   24) "slave"
   25) "role-reported-time"
   26) "286309"
   27) "master-link-down-time"
   28) "1482321036000"
   29) "master-link-status"
   30) "err"
   31) "master-host"
   32) "192.168.126.133"
   33) "master-port"
   34) "6379"
   35) "slave-priority"
   36) "100"
   37) "slave-repl-offset"
   38) "1"
2)  1) "name"
    2) "192.168.126.138:6379"
    3) "ip"
    4) "192.168.126.138"
    5) "port"
    6) "6379"
    7) "runid"
    8) "0397cb460b84ca7df4650483ae8707bda53c2ddc"
    9) "flags"
   10) "slave"
   11) "pending-commands"
   12) "0"
   13) "last-ping-sent"
   14) "0"
   15) "last-ok-ping-reply"
   16) "582"
   17) "last-ping-reply"
   18) "582"
   19) "down-after-milliseconds"
   20) "30000"
   21) "info-refresh"
   22) "5661"
   23) "role-reported"
   24) "slave"
   25) "role-reported-time"
   26) "568112"
   27) "master-link-down-time"
   28) "0"
   29) "master-link-status"
   30) "ok"
   31) "master-host"
   32) "192.168.126.133"
   33) "master-port"
   34) "6379"
   35) "slave-priority"
   36) "100"
   37) "slave-repl-offset"
   38) "117625"
127.0.0.1:26379> 

可以看到133被选举为master，128与138为slave

查看138的redis配置文件redis.conf,Master已经被修改

slaveof 192.168.126.133 6379

查看138的redis配置文件sentinel已经被修改

sentinel config-epoch zabbix 1
sentinel leader-epoch zabbix 1
sentinel known-slave zabbix 192.168.126.138 6379
# Generated by CONFIG REWRITE
sentinel known-slave zabbix 192.168.126.128 6379
sentinel known-sentinel zabbix 192.168.126.133 26379 17607179ad13102b1561dcd6a999cff732ac0f98
sentinel known-sentinel zabbix 192.168.126.128 26379 b4354571b53ffdbc7369913f3c2c87a31f2a22a4
sentinel current-epoch 1

查看128与133的配置文件，也相应的修改。
133的slaveof配置被擦除

总结：
[color=green]redis的主从与mysql的主从有点类似，都要配置slave的Master，sentinel类与Hadoop HA环境下的Zookeeper的作用，当集群中master宕机时，sentinel通过GrossIP协议选择出Master，并修改相应的redis.conf,与sentinel.conf配置文件。
[/color]