搭建高可用及负载均衡的REDIS

精选转载

jbaowei2000 2016-09-22 17:11:24 博主文章分类：redis

文章标签 搭建高可用及负载均衡的REDIS 文章分类 Redis 数据库

测试环境

机器
- 母机：Ubuntu 4核 16G
- 虚拟机:centos6.5-64 单核 1G 独立ip 3个
  虚拟机使用virtual box,centos为64位6.5.具体信息如下
```
192.168.0.231   Redis-master192.168.0.231   Redis-slave192.168.0.239   Virtual IP
```
软件
- Reidis
- Keepalived

安装前配置

虚拟机最小化安装，更新系统及安装必须组件

$ yum install -y openssl openssl-devel kernel kernel-devel gcc wget tcl libnl-devel$ yum upgrade

$ reboot  //重启系统

安装Keepalived

安装keepalived 下载地址:http://www.keepalived.org/software/keepalived-1.2.13.tar.gz 安装：

$ wget http://www.keepalived.org/software/keepalived-1.2.13.tar.gz$ tar -zxvf keepalived-1.2.13.tar.gz$ cd keepalived-1.2.13$ ln -s /usr/src/kernels/`uname -r`-`uname -m`/ /usr/src/linux$ ./configure  --prefix=/usr/local/keepalived  --sysconf=/etc$ chkconfig --add keepalived//查看是否成功$ service keepalived status

安装redis

下载Redis,目前Redis的稳定版本是2.8.17，下载地址Redis-2.8.17

$ wget http://download.redis.io/releases/redis-2.8.17.tar.gz$ tar -zxvf redis-2.8.17.tar.gz

$ cd redis-2.8.17$ make && make test

$ mkdir -p /usr/local/redis/{bin,var,etc}

$ cp ./src/{redis-benchmark,redis-cli,redis-server,redis-check-dump,redis-check-aof} /usr/local/redis/bin/$ cp /usr/local/redis/bin/* /usr/local/bin/

配置

Master配置

先配置好Redis，主从复制成功之后再行配置Keepalived。

Master-Redis配置

$ vim /usr/local/redis/etc/redis.conf

修改如下值

daemonize yes
pidfile /var/run/redis.pid
port 6379tcp-backlog 511timeout 0tcp-keepalive 0loglevel notice
logfile "/usr/local/redis/log/redis.log"databases 16save 900 1save 300 10save 60 10000stop-writes-on-bgsave-error yes
rdbcompression yes
rdbchecksum yes
dbfilename dump.rdb
dir /usr/local/redis/dataslave-serve-stale-data yes
slave-read-only no
repl-disable-tcp-nodelay no
slave-priority 100appendonly yes
appendfilename "appendonly.aof"appendfsync everysec
no-appendfsync-on-rewrite no
auto-aof-rewrite-percentage 100auto-aof-rewrite-min-size 64mb
lua-time-limit 5000slowlog-log-slower-than 10000slowlog-max-len 128notify-keyspace-events ""hash-max-ziplist-entries 512hash-max-ziplist-value 64list-max-ziplist-entries 512list-max-ziplist-value 64set-max-intset-entries 512zset-max-ziplist-entries 128zset-max-ziplist-value 64activerehashing yes
client-output-buffer-limit normal 0 0 0client-output-buffer-limit slave 256mb 64mb 60client-output-buffer-limit pubsub 32mb 8mb 60hz 10aof-rewrite-incremental-fsync yes

Master-Keepalived配置

global_defs {
    router_id redis
}
vrrp_script chk_redis {
    script "/etc/keepalived/scripts/redis_check.sh"   ###监控脚本
    interval 2                  ###监控时间
    timeout  2                                  ###超时时间
    fall     3                  ###重试次数}
vrrp_instance redis {
  state MASTER                  ###设置为MASTER
  interface eth0              ###监控网卡  virtual_router_id 52  priority 101              ###权重值  authentication {
         auth_type PASS          ###加密
         auth_pass redis        ###密码
  }
  track_script {
    chk_redis              ###执行上面定义的chk_redis
  }
  virtual_ipaddress {       192.168.0.239            ###VIP
  }
  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/redsi_stop.sh
}

也可在本地编辑好后使用scp上传到服务器。 Master-keepalived脚本redis_master.sh:

  #!/bin/bash
  REDISCLI="/usr/local/redis/bin/redis-cli"
  LOGFILE="/usr/local/redis/var/keepalived-redis-state.log"
  echo "[master]" >> $LOGFILE
  date >> $LOGFILE
  echo "Being master...." >> $LOGFILE 2>&1
  echo "Run SLAVEOF cmd ..." >> $LOGFILE
  $REDISCLI SLAVEOF 192.168.0.232 6379 >> $LOGFILE  2>&1
  sleep 10 #延迟10秒以后待数据同步完成后再取消同步状态
  echo "Run SLAVEOF NO ONE cmd ..." >> $LOGFILE
  $REDISCLI SLAVEOF NO ONE >> $LOGFILE 2>&1

Master-keepalived脚本redis_backup.sh:

  #!/bin/bash
  REDISCLI="/usr/local/redis/bin/redis-cli"
  LOGFILE="/usr/local/redis/log/keepalived-redis-state.log"
  echo "[backup]" >> $LOGFILE
  date >> $LOGFILE
  echo "Being slave...." >> $LOGFILE 2>&1
  sleep 15 #延迟15秒待数据被对方同步完成之后再切换主从角色
  echo "Run SLAVEOF cmd ..." >> $LOGFILE
  $REDISCLI SLAVEOF 192.168.0.232 6379 >> $LOGFILE  2>&1

Backup配置

同样先配置好Redis，主从复制成功之后再行配置Keepalived。

Backup-Redis配置

$ vim /usr/local/redis/etc/redis.conf

修改如下值

daemonize yes
pidfile /var/run/redis.pid
port 6379tcp-backlog 511timeout 0tcp-keepalive 0loglevel notice
logfile "/usr/local/redis/log/redis.log"databases 16save 900 1save 300 10save 60 10000stop-writes-on-bgsave-error yes
rdbcompression yes
rdbchecksum yes
dbfilename dump.rdb
dir /usr/local/redis/dataslave-serve-stale-data yes
slave-read-only no
repl-disable-tcp-nodelay no
slave-priority 100appendonly yes
appendfilename "appendonly.aof"appendfsync everysec
no-appendfsync-on-rewrite no
auto-aof-rewrite-percentage 100auto-aof-rewrite-min-size 64mb
lua-time-limit 5000slowlog-log-slower-than 10000slowlog-max-len 128notify-keyspace-events ""hash-max-ziplist-entries 512hash-max-ziplist-value 64list-max-ziplist-entries 512list-max-ziplist-value 64set-max-intset-entries 512zset-max-ziplist-entries 128zset-max-ziplist-value 64activerehashing yes
client-output-buffer-limit normal 0 0 0client-output-buffer-limit slave 256mb 64mb 60client-output-buffer-limit pubsub 32mb 8mb 60hz 10aof-rewrite-incremental-fsync yes
SLAVEOF 192.168.0.231 6379

Backup-Keepalived配置

global_defs {
    router_id redis
}
vrrp_script chk_redis {
    script "/etc/keepalived/scripts/redis_check.sh"   ###监控脚本
    interval 2                  ###监控时间
    timeout  2                                  ###超时时间
    fall     3                  ###重试次数}
vrrp_instance redis {
  state BACKUP                  ###设置为MASTER
  interface eth0              ###监控网卡  virtual_router_id 51  priority 100              ###权重值  authentication {
         auth_type PASS          ###加密
         auth_pass redis        ###密码
  }
  track_script {
    chk_redis              ###执行上面定义的chk_redis
  }
  virtual_ipaddress {       192.168.0.239            ###VIP
  }
  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/redsi_stop.sh
}

也可在本地编辑好后使用scp上传到服务器。 Master-keepalived脚本redis_master.sh:

  #!/bin/bash
  REDISCLI="/usr/local/redis/bin/redis-cli"
  LOGFILE="/usr/local/redis/var/keepalived-redis-state.log"
  echo "[master]" >> $LOGFILE
  date >> $LOGFILE
  echo "Being master...." >> $LOGFILE 2>&1
  echo "Run SLAVEOF cmd ..." >> $LOGFILE
  $REDISCLI SLAVEOF 192.168.0.231 6379 >> $LOGFILE  2>&1
  sleep 10 #延迟10秒以后待数据同步完成后再取消同步状态
  echo "Run SLAVEOF NO ONE cmd ..." >> $LOGFILE
  $REDISCLI SLAVEOF NO ONE >> $LOGFILE 2>&1

Master-keepalived脚本redis_backup.sh:

  #!/bin/bash
  REDISCLI="/usr/local/redis/bin/redis-cli"
  LOGFILE="/usr/local/redis/log/keepalived-redis-state.log"
  echo "[backup]" >> $LOGFILE
  date >> $LOGFILE
  echo "Being slave...." >> $LOGFILE 2>&1
  sleep 15 #延迟15秒待数据被对方同步完成之后再切换主从角色
  echo "Run SLAVEOF cmd ..." >> $LOGFILE
  $REDISCLI SLAVEOF 192.168.0.231 6379 >> $LOGFILE  2>&1

公共脚本配置

redis_chekc.sh
可以使用netcat来检测redis服务的状态.
安装nc，使用nc检测redis的服务而不是ping
$ wget http://sourceforge.net/projects/netcat/files/netcat/0.7.1/netcat-0.7.1.tar.gz
$ tar -zxvf netcat-0.7.1.tar.gz
$ ./configure
$ make && make install
建立脚本:
```
$ vim /etc/keepalived/scripts/redis_check.sh
```
脚本内容:
```
#!/bin/bashALIVE=`/usr/local/redis/bin/redis-cli PING`if ["$ALIVE" == "PONG"] ;then
      echo $ALIVE
      exit 0else 
      echo $ALIVE
      exit 1fi
```

redis_fault.sh

建立脚本:

$ vim /etc/keepalived/scripts/redis_fault.sh

脚本内容:

# !/bin/bashLOGFILE=/usr/local/redis/var/keepalived-redis-state.log 
echo "[faule]" >> $LOGFILE date >> $LOGFILE

redis_stop.sh

建立脚本:

$ vim /etc/keepalived/scripts/redis_stop.sh

脚本内容:

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

测试

主从复制测试
先打开Master上的Redis，再打开Backup上的Redis，分别查看info
```
$ redis-cli info
```
Master的rule显示master为正常，slave会显示为192.168.0.232
在Master上写入数据，在Backup读取
```
$ redis-cli set a hello  //Master 写入数据$ redis-cli get a        //Backup 读取数据"hello"
```
主从复制正常。
主从切换测试
先后启动Master和Backup上的keepalived，在Master上查看vip
```
$ ip a
```
从图中可以看出vip已经绑定在Master上，现在通过vip来进行数据测试
```
$ redis-cli -h 192.168.0.239 set a test
$ redis-cli -h 192.168.0.239 get a //测试虚拟ip数据读取//在主机和从机上分别进行数据读取$ redis-cli get a
```
从图中可以看出测试无问题。
现在关闭Master上的redis，来看VIP能否正常漂移，并进行数据测试。
```
$ killall -9 redis-server$ redis-cli -h 192.168.0.232 info
```
现在给redis写入数据，测试看Master恢复服务后能否正常恢复数据
```
$ redis-cli -h 192.168.0.239 set a nihao
```
恢复Master上的Redis，查看ip a
```
$ service redis-server start$ ip a$ redis-cli get a
```
数据恢复，VIP切换回Master，Redis高可用环境搭建成功.