Linux下 Redis主从架构持久化操作详述

Redis主从架构持久化存在一个问题，即前次测试的结论，持久化需要配置在主实例上才能跨越实例保证数据不丢失，这样以来主实例在持久化数据到硬 盘的过程中，势必会造成磁盘的I/O等待，经过实际测试，这个持久化写硬盘的过程给应用程序带来的影响无法忍受；因而在大多数场景下，会考虑把持久化配置 在从实例上，当主实例宕机后，通过手动或者自动的方式将从实例提升为主实例，继续提供服务！当主实例恢复后，先从原从实例上同步数据，同步完成后再恢复到 原始的主从状态！要实现这种的要求，需要有keepalive的配合，一方面keepalive提供了VIP，可以避免修改应用程序连接，同时redis 实例的配置文件监听部分也需要修改为全网监听；另一方面keepalive定时调度脚本来监控主从实例的状态，根据具体情况进行切换！本文将重点介绍下使 用keepalive实现redis主从自动failover！

环境介绍
 操作系统版本均为：rhel5.4 64bit
 redis版本：2.6.4
 redis实例端口均为：6379
 redis实例密码均为：123
 VIP：192.168.1.120
主实例为server11(192.168.1.112)
从实例为server12(192.168.1.113，开启快照持久化)

一：安装keepalive软件,server11安装完成后直接scp至server12上即可

    [root@server11 ~]# wget http://keepalived.org/software/keepalived-1.1.19.tar.gz 
    [root@server11 ~]# tar -zxvf ../tarbag/keepalived-1.1.19.tar.gz 
    [root@server11 ~]# cd keepalived-1.1.19/ 
    [root@server11 ~]# ./configure --prefix=/usr/local/keepalived && make && make install

二：配置主节点server11配置文件

    [root@server11 ~]# cat /usr/local/keepalived/etc/keepalived/keepalived.conf 
    ! Configuration File for keepalived 
   
    global_defs { 
    router_id LVS_DEVEL 
    } 
   
    vrrp_script Monitor_redis { 
    script "/usr/local/scripts/redis_monitor.sh" 
    interval 2 
    weight 2   
    } 
   
    vrrp_instance VI_1{ 
    state MASTER 
    interface eth0 
    virtual_router_id 51 
    mcast_src_ip 192.168.1.112 
    priority  100 
    advert_int 1 
    authentication { 
    auth_type PASS 
    auth_pass password_123 
    } 
    track_script { 
    Monitor_redis 
    } 
    virtual_ipaddress { 
    192.168.1.120 
    } 
    notify_fault  /usr/local/scripts/redis_fault.sh   
    notify_stop  /usr/local/scripts/redis_stop.sh   
   
    }

三：配置从节点server12配置文件

    [root@server12 ~]# cat /usr/local/keepalived/etc/keepalived/keepalived.conf 
    ! Configuration File for keepalived 
   
    global_defs { 
    router_id LVS_DEVEL 
    } 
   
    vrrp_script Monitor_redis { 
    script "/usr/local/scripts/redis_monitor.sh" 
    interval 2 
    weight 2   
    } 
   
    vrrp_instance VI_1{ 
    state BACKUP 
    interface eth0 
    virtual_router_id 51 
    mcast_src_ip 192.168.1.113 
    priority  99 
    advert_int 1 
    authentication { 
    auth_type PASS 
    auth_pass password_123 
    } 
    track_script { 
    Monitor_redis 
    } 
    virtual_ipaddress { 
    192.168.1.120 
    } 
    notify_master /usr/local/scripts/redis_master.sh   
    notify_backup /usr/local/scripts/redis_backup.sh   
    notify_fault  /usr/local/scripts/redis_fault.sh   
    notify_stop  /usr/local/scripts/redis_stop.sh   
   
    }

四：准备相关的脚本，主从实例上都需要存在这些脚本，同时注意脚本需要由可执行权限

    [root@server11 ~]# cat /usr/local/scripts/redis_monitor.sh 
    #!/bin/bash   
    ALIVE=$(/usr/local/redis2/bin/redis-cli -h 192.168.1.112 -p 6379 -a 123 PING) 
   
    if [ "$ALIVE" == "PONG" ]; then 
        echo $ALIVE   
        exit 0   
        else 
        echo $ALIVE   
        killall -9 keepalived 
        service network restart 
        exit 1   
    fi 
   
    [root@server11 ~]# sh /usr/local/scripts/redis_monitor.sh 
    PONG 
   
    [root@server11 ~]# cat /usr/local/scripts/redis_master.sh   
    #!/bin/bash   
    REDISCLI="/usr/local/redis2/bin/redis-cli -h 192.168.1.112 -p 6379 -a 123" 
    LOGFILE="/usr/local/redis2/var/keepalived-redis-state.log" 
   
    echo "[master]" >> $LOGFILE   
    date >> $LOGFILE   
    echo "Being master...." >> $LOGFILE 2>&1   
    echo "Run SLAVEOF cmd ..." >> $LOGFILE   
    REDISCLI SLAVEOF 192.168.1.113 6379 >>LOGFILE  2>&1   
    sleep 10 
    echo "Run SLAVEOF NO ONE cmd ..." >> $LOGFILE   
    REDISCLI SLAVEOF NO ONE >>LOGFILE 2>&1 
   
   
    [root@server11 ~]# cat /usr/local/scripts/redis_backup.sh 
    #!/bin/bash   
    REDISCLI="/usr/local/redis2/bin/redis-cli -h 192.168.1.112 -p 6379 -a 123" 
    LOGFILE="/usr/local/redis2/var/keepalived-redis-state.log" 
   
    echo "[backup]" >> $LOGFILE   
    date >> $LOGFILE   
    echo "Being slave...." >> $LOGFILE 2>&1   
    sleep 15   
    echo "Run SLAVEOF cmd ..." >> $LOGFILE   
    REDISCLI SLAVEOF 192.168.1.113 6379 >>LOGFILE  2>&1 
   
    [root@server11 ~]# cat /usr/local/scripts/redis_stop.sh 
    #!/bin/bash 
    LOGFILE="/usr/local/redis2/var/keepalived-redis-state.log" 
    echo "[stop]" >> $LOGFILE 
    date >> $LOGFILE 
    [root@server11 ~]# cat /usr/local/scripts/redis_fault.sh
    #!/bin/bash 
    LOGFILE="/usr/local/redis2/var/keepalived-redis-state.log"
    echo "[fault]" >> $LOGFILE 
    date >> $LOGFILE

五：主从实例分别启动keepalive进程，测试VIP是否正常(这里就要修改redis配置文件的监听地址为0.0.0.0)

[root@server11 ~]# /usr/local/keepalived/sbin/keepalived -D -f  /usr/local/keepalived/etc/keepalived/keepalived.conf 
[root@server11 ~]# tail -f /var/log/messages 
Dec 12 09:25:49 server11 Keepalived_healthcheckers[7710]: Configuration is using : 5499 Bytes 
Dec 12 09:25:49 server11 Keepalived_healthcheckers[7710]: Using LinkWatch kernel netlink reflector... 
Dec 12 09:25:49 server11 Keepalived_vrrp[7712]: VRRP sockpool: [ifindex(2), proto(112), fd(12,13)] 
Dec 12 09:25:49 server11 Keepalived_vrrp[7712]: VRRP_Script(Monitor_redis) succeeded 
Dec 12 09:25:50 server11 Keepalived_vrrp[7712]: VRRP_Instance(VI_1{) Transition to MASTER STATE 
Dec 12 09:25:51 server11 Keepalived_vrrp[7712]: VRRP_Instance(VI_1{) Entering MASTER STATE 
Dec 12 09:25:51 server11 Keepalived_vrrp[7712]: VRRP_Instance(VI_1{) setting protocol VIPs. 
Dec 12 09:25:51 server11 Keepalived_vrrp[7712]: VRRP_Instance(VI_1{) Sending gratuitous ARPs on eth0 for 192.168.1.120 
Dec 12 09:25:51 server11 avahi-daemon[4519]: Registering new address record for 192.168.1.120 on eth0. 
Dec 12 09:25:51 server11 Keepalived_healthcheckers[7710]: Netlink reflector reports IP 192.168.1.120 added 
Dec 12 09:25:51 server11 Keepalived_vrrp[7712]: Netlink reflector reports IP 192.168.1.120 added 
Dec 12 09:25:56 server11 Keepalived_vrrp[7712]: VRRP_Instance(VI_1{) Sending gratuitous ARPs on eth0 for 192.168.1.120 
 
[root@server11 ~]# ip a |grep 192 
    inet 192.168.1.112/24 brd 192.168.1.255 scope global eth0 
    inet 192.168.1.120/32 scope global eth0 
 
[root@server12 ~]# /usr/local/keepalived/sbin/keepalived -D -f /usr/local/keepalived/etc/keepalived/keepalived.conf 
[root@server12 ~]# tail -f /var/log/messages 
Dec 12 09:26:55 server12 Keepalived_healthcheckers[3106]: Configuration is using : 5595 Bytes 
Dec 12 09:26:55 server12 Keepalived_vrrp[3108]: VRRP_Instance(VI_1{) Entering BACKUP STATE 
Dec 12 09:26:55 server12 Keepalived_healthcheckers[3106]: Using LinkWatch kernel netlink reflector... 
Dec 12 09:26:55 server12 Keepalived_vrrp[3108]: VRRP sockpool: [ifindex(2), proto(112), fd(12,13)] 
Dec 12 09:26:55 server12 Keepalived_vrrp[3108]: VRRP_Script(Monitor_redis) succeeded 
 
[root@server11 ~]# /usr/local/redis2/bin/redis-cli -h 192.168.1.120 -a 123 info |grep -A 3 'Replication' 
# Replication 
role:master 
connected_slaves:1 
slave0:192.168.1.113,6379,online

六：主实例写入测试数据，该脚本原则上会写入25条测试数据，不过由于未优化redis默认并发数，会导致一些写入请求失败，最终功写入231839条测 试数据，占内存总大小为25M左右，写入过程中可以观察主从实例的持久化文件变化情况，主实例的持久化文件维持在30k，从实例的则不断的扩展！

    [root@server11 ~]# cat test.sh 
    #!/bin/bash   
    REDISCLI="/usr/local/redis2/bin/redis-cli -h 192.168.1.120 -a 123 -n 1 SET" 
    ID=1   
    while(($ID<50001))   
    do 
    INSTANCE_NAME="i-2-$ID-VM" 
    UUID=`cat /proc/sys/kernel/random/uuid`   
    PRIVATE_IP_ADDRESS=10.`echo "RANDOM % 255 + 1" | bc`.`echo "RANDOM % 255 + 1" | bc`.`echo 
   
    "$RANDOM % 255 + 1" | bc`\   
    CREATED=`date "+%Y-%m-%d %H:%M:%S"`   
    REDISCLI vm_instance:ID:instance_name "$INSTANCE_NAME" 
    REDISCLI vm_instance:ID:uuid "$UUID" 
    REDISCLI vm_instance:ID:private_ip_address "$PRIVATE_IP_ADDRESS" 
    REDISCLI vm_instance:ID:created "$CREATED" 
    REDISCLI vm_instance:INSTANCE_NAME:id "$ID" 
    ID=((ID+1))   
    done 
   
    [root@server11 ~]# sh test.sh 
    [root@server11 redis2]# /usr/local/redis2/bin/redis-cli -h 192.168.1.120 -a 123 info |egrep 
   
    'used_memory_peak_human|db1:keys' 
    used_memory_peak_human:24.98M 
    db1:keys=231839,expires=0

七：模拟主实例故障，观察日志输出，验证从实例是否能成功接管VIP，同时将实例变成读写模式

    [root@server11 ~]# killall -9  redis-server 
    [root@server11 ~]# ip a |grep 192 
        inet 192.168.1.112/24 brd 192.168.1.255 scope global eth0 
   
    [root@server11 ~]# ps -ef |grep redis 
    root    15886  6458  0 09:49 pts/0    00:00:00 grep redis 
    [root@server11 ~]# ps -ef |grep keep 
    root    16029  6458  0 09:49 pts/0    00:00:00 grep keep 
   
   
    [root@server12 ~]# tail -f /usr/local/redis2/var/keepalived-redis-state.log 
    [master] 
    Wed Dec 12 09:48:52 CST 2012 
    Being master.... 
    Run SLAVEOF cmd ... 
    OK Already connected to specified master 
    Run SLAVEOF NO ONE cmd ... 
    OK 
   
    [root@server12 ~]# tail -f /var/log/messages 
    Dec 12 09:48:51 server12 Keepalived_vrrp[3108]: VRRP_Instance(VI_1{) Transition to MASTER STATE 
    Dec 12 09:48:52 server12 Keepalived_vrrp[3108]: VRRP_Instance(VI_1{) Entering MASTER STATE 
    Dec 12 09:48:52 server12 Keepalived_vrrp[3108]: VRRP_Instance(VI_1{) setting protocol VIPs. 
    Dec 12 09:48:52 server12 Keepalived_vrrp[3108]: VRRP_Instance(VI_1{) Sending gratuitous ARPs on eth0 for 192.168.1.120 
    Dec 12 09:48:52 server12 Keepalived_vrrp[3108]: Netlink reflector reports IP 192.168.1.120 added 
    Dec 12 09:48:52 server12 avahi-daemon[2921]: Registering new address record for 192.168.1.120 on eth0. 
    Dec 12 09:48:52 server12 Keepalived_healthcheckers[3106]: Netlink reflector reports IP 192.168.1.120 added 
    Dec 12 09:48:57 server12 Keepalived_vrrp[3108]: VRRP_Instance(VI_1{) Sending gratuitous ARPs on eth0 for 192.168.1.120 
   
    [root@server12 ~]# ip a |grep 192 
        inet 192.168.1.113/24 brd 192.168.1.255 scope global eth0 
        inet 192.168.1.120/32 scope global eth0 
   
    [root@server12 ~]# /usr/local/redis2/bin/redis-cli -h 192.168.1.120 -a 123 info |grep -A 3 'Replication' 
    # Replication 
    role:master 
    connected_slaves:0

    [root@server12 ~]# sh test.sh 
    [root@server12 ~]# /usr/local/redis2/bin/redis-cli -h 192.168.1.120 -a 123 info |egrep 
   
    'used_memory_peak_human|db1:keys' 
    used_memory_peak_human:26.78M 
    db1:keys=249925,expires=0

九:主实例角色的恢复过程，使用shell脚本自动恢复

    [root@server11 ~]# ssh-keygen 
    [root@server11 ~]# cd .ssh/ 
    [root@server11 .ssh]# ssh-copy-id -i id_rsa.pub [email protected] 
    [root@server11 ~]# cat /usr/local/scripts/recover_mastart.sh 
    #!/bin/sh 
    ALIVE=$(/usr/local/redis2/bin/redis-cli -h 192.168.1.113 -p 6379 -a 123 PING) 
    MDB=/usr/local/redis2/master_dump.rdb 
    SDB=/usr/local/redis2/slave_dump.rdb 
   
    if [ "$ALIVE" == "PONG" ]; then 
        echo $ALIVE   
        scp [email protected]:$SDB  $MDB 
        else 
        echo $ALIVE 
        exit 1   
    fi 
   
    /usr/local/redis2/bin/redis-server /usr/local/redis2/etc/redis.conf 
    /usr/local/keepalived/sbin/keepalived -D -f   
   
    /usr/local/keepalived/etc/keepalived/keepalived.conf 
   
    [root@server11 ~]# chmod +x  /usr/local/scripts/recover_mastart.sh 
    [root@server11 ~]# sh /usr/local/scripts/recover_mastart.sh 

十：验证数据完整性和主从角色恢复情况

    [root@server11 ~]# /usr/local/redis2/bin/redis-cli -h 192.168.1.120 -a 123 info |egrep 'used_memory_peak_human|db1:keys' 
    used_memory_peak_human:26.78M 
    db1:keys=249925,expires=0
   
    [root@server11 ~]#  /usr/local/redis2/bin/redis-cli -h 192.168.1.120 -a 123 info |grep -A 3 'Replication' 
    # Replication 
    role:master 
    connected_slaves:1 
    slave0:192.168.1.113,6379,online 
   
    [root@server12 ~]#  /usr/local/redis2/bin/redis-cli -h 192.168.1.113 -a 123 info |grep -A 3 'Replication' 
    # Replication 
    role:slave 
    master_host:192.168.1.112 
    master_port:6379 
   
    [root@server12 ~]# /usr/local/redis2/bin/redis-cli -h 192.168.1.120 -a 123 info |egrep 'used_memory_peak_human|db1:keys' 
    used_memory_peak_human:26.78M 
    db1:keys=249925,expires=0
   
    主实例keepalive日志： 
    [root@server11 ~]# tail -f /var/log/messages 
    Dec 12 10:08:13 server11 Keepalived_vrrp[20231]: VRRP sockpool: [ifindex(2), proto(112), fd(11,12)] 
    Dec 12 10:08:13 server11 Keepalived_vrrp[20231]: VRRP_Script(Monitor_redis) succeeded 
    Dec 12 10:08:13 server11 Keepalived_vrrp[20231]: VRRP_Instance(VI_1{) Transition to MASTER STATE 
    Dec 12 10:08:13 server11 Keepalived_vrrp[20231]: VRRP_Instance(VI_1{) Received higher prio advert 
    Dec 12 10:08:13 server11 Keepalived_vrrp[20231]: VRRP_Instance(VI_1{) Entering BACKUP STATE 
    Dec 12 10:08:15 server11 Keepalived_vrrp[20231]: VRRP_Instance(VI_1{) forcing a new MASTER election 
    Dec 12 10:08:16 server11 Keepalived_vrrp[20231]: VRRP_Instance(VI_1{) Transition to MASTER STATE 
    Dec 12 10:08:17 server11 Keepalived_vrrp[20231]: VRRP_Instance(VI_1{) Entering MASTER STATE 
    Dec 12 10:08:17 server11 Keepalived_vrrp[20231]: VRRP_Instance(VI_1{) setting protocol VIPs. 
    Dec 12 10:08:17 server11 Keepalived_healthcheckers[20230]: Netlink reflector reports IP 192.168.1.120 added 
    Dec 12 10:08:17 server11 Keepalived_vrrp[20231]: VRRP_Instance(VI_1{) Sending gratuitous ARPs on eth0 for 192.168.1.120 
    Dec 12 10:08:17 server11 Keepalived_vrrp[20231]: Netlink reflector reports IP 192.168.1.120 added 
    Dec 12 10:08:17 server11 avahi-daemon[4519]: Registering new address record for 192.168.1.120 on eth0. 
   
    [root@server11 ~]# ip a |grep 192 
        inet 192.168.1.112/24 brd 192.168.1.255 scope global eth0 
        inet 192.168.1.120/32 scope global eth0 
   
    从实例keepalive日志： 
    [root@server12 ~]# tail -f /var/log/messages 
    Dec 12 09:56:01 server12 last message repeated 4 times 
    Dec 12 10:08:13 server12 Keepalived_vrrp[3108]: VRRP_Instance(VI_1{) Received lower prio advert, forcing new election 
    Dec 12 10:08:13 server12 Keepalived_vrrp[3108]: VRRP_Instance(VI_1{) Sending gratuitous ARPs on eth0 for 192.168.1.120 
    Dec 12 10:08:15 server12 Keepalived_vrrp[3108]: VRRP_Instance(VI_1{) Received higher prio advert 
    Dec 12 10:08:15 server12 Keepalived_vrrp[3108]: VRRP_Instance(VI_1{) Entering BACKUP STATE 
    Dec 12 10:08:15 server12 Keepalived_vrrp[3108]: VRRP_Instance(VI_1{) removing protocol VIPs. 
    Dec 12 10:08:15 server12 Keepalived_healthcheckers[3106]: Netlink reflector reports IP 192.168.1.120 removed 
    Dec 12 10:08:15 server12 Keepalived_vrrp[3108]: Netlink reflector reports IP 192.168.1.120 removed 
    Dec 12 10:08:15 server12 avahi-daemon[2921]: Withdrawing address record for 192.168.1.120 on eth0. 
   
    从实例角色转换日志： 
    [root@server12 ~]# tail -f /usr/local/redis2/var/keepalived-redis-state.log 
    [backup] 
    Wed Dec 12 10:08:15 CST 2012 
    Being slave.... 
    Run SLAVEOF cmd ... 
    OK

下面关于