1. 前言

　　负载均衡LB，高可用HA，这一小结主要讲双机热备方案保证高可用。这里选择Keepalived作为双机热备方案，下面就对具体的配置进行了解。

2. 下载Keepalived

　　wget http://www.keepalived.org/software/keepalived-1.4.0.tar.gz

　　文档 http://www.keepalived.org/doc

　　参考 https://www.cnblogs.com/abclife/p/7909818.html

　　https://www.cnblogs.com/kevingrace/p/6138185.html

　　系统 Debian 8

1 ./configure --prefix=/opt/keepalive #这一步，可能要额外安装一些依赖

1 make2 make install3 mkdir /etc/keepalived4 cp ./etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf5 cp ./sbin/keepalived /usr/sbin/6 vim /etc/init.d/keepalived7 chmod a+x /etc/init.d/keepalived8 service keepalived start

　　/etc/init.d/keepalived 加入如下内容

1 #!/bin/sh   2 #   3 # keepalived   High Availability monitor built upon LVS and VRRP   4 #   5 # chkconfig:   - 86 14   6 # description: Robust keepalive facility to the Linux Virtual Server project \   7 #              with multilayer TCP/IP stack checks.   8      9 ### BEGIN INIT INFO  10 # Provides: keepalived  11 # Required-Start: $local_fs $network $named $syslog  12 # Required-Stop: $local_fs $network $named $syslog  13 # Should-Start: smtpdaemon httpd  14 # Should-Stop: smtpdaemon httpd  15 # Default-Start:   16 # Default-Stop: 0 1 2 3 4 5 6  17 # Short-Description: High Availability monitor built upon LVS and VRRP  18 # Description:       Robust keepalive facility to the Linux Virtual Server  19 #                    project with multilayer TCP/IP stack checks.  20 ### END INIT INFO  21     22 # Source function library.  23 . /etc/rc.d/init.d/functions  24     25 exec="/usr/sbin/keepalived"  26 prog="keepalived"  27 config="/etc/keepalived/keepalived.conf"  28     29 [ -e /etc/sysconfig/$prog ] && . /etc/sysconfig/$prog  30     31 lockfile=/var/lock/subsys/keepalived  32     33 start() {  34     [ -x $exec ] || exit 5  35     [ -e $config ] || exit 6  36     echo -n $"Starting $prog: "  37     daemon $exec $KEEPALIVED_OPTIONS  38     retval=$?  39     echo  40     [ $retval -eq 0 ] && touch $lockfile  41     return $retval  42 }  43     44 stop() {  45     echo -n $"Stopping $prog: "  46     killproc $prog  47     retval=$?  48     echo  49     [ $retval -eq 0 ] && rm -f $lockfile  50     return $retval  51 }  52     53 restart() {  54     stop  55     start  56 }  57     58 reload() {  59     echo -n $"Reloading $prog: "  60     killproc $prog -1  61     retval=$?  62     echo  63     return $retval  64 }  65     66 force_reload() {  67     restart  68 }  69     70 rh_status() {  71     status $prog  72 }  73     74 rh_status_q() {  75     rh_status &>/dev/null  76 }  77     78     79 case "$1" in  80     start)  81         rh_status_q && exit 0  82         $1  83         ;;  84     stop)  85         rh_status_q || exit 0  86         $1  87         ;;  88     restart)  89         $1  90         ;;  91     reload)  92         rh_status_q || exit 7  93         $1  94         ;;  95     force-reload)  96         force_reload  97         ;;  98     status)  99         rh_status 100         ;; 101     condrestart|try-restart) 102         rh_status_q || exit 0 103         restart 104         ;; 105     *) 106         echo $"Usage: $0 {start|stop|status|restart|condrestart|try-restart|reload|force-reload}" 107         exit 2 108 esac 109 exit $?

3. 双机热备（主从模式）

　　修改配置文件 keepalived.conf

　　vim /etc/keeplived/keeplived.conf

 

1 global_defs { 2     notification_email { #指定Keepalived在发生事情的时候，发送邮件通知，每行一个地址 3         yyyyy@qq.com 4 　　} 5     notification_email_from yyyyy@qq.com #指定发件人 6     smtp_server 192.168.8.208 #发送email的smtp地址 7     smtp_connect_timeout 30 #超时时间 8     router_id nginx_dev #运行Keepalived的机器标识号，可以相同也可以不同 9 }10 11 vrrp_instance nginx_dev {12     state MASTER13     interface eth114     #mcast_src_ip 172.16.23.20315     virtual_router_id 10016     priority 20017     advert_int 518     authentication {19         auth_type PASS20         auth_pass 12345621     }22     virtual_ipaddress {23         172.16.23.22224     }25 }

 

　　然后 service keepalived restart

　　然后就可以通过 172.16.23.222 访问当前主机了。如果访问不了的要判断是否所处于同个网络中，即可以通过路由访问到的。这个Keepalived就是通过发广播包这里的通知内网机器实现的。

　　同理在另外一台电脑上配置上面信息 把state MASTER 改为 state BACKUP

4. 双机热备（主主模式）

　　主从模式的话，由于一般情况下是不会出现宕机，所以往往会有一台机器浪费，这样是对机器的浪费，所以现在双机热备主主模式是比较推荐的。所谓的主主模式，就是建立两个实例，互为主从而已。

　　172.16.23.203 配置如下

1 global_defs { 2     notification_email {  #指定Keepalived在发生事情的时候，发送邮件通知，每行一个地址 3         xxx@aa.com 4     } 5     notification_email_from xxx@aa.com   #指定发件人 6     smtp_server 192.168.8.208   #发送email的smtp地址 7     smtp_connect_timeout 30     #超时时间 8     router_id nginx_dev_1       #运行Keepalived的机器标识号，可以相同也可以不同 9     router_id nginx_dev_210 }11 12 vrrp_instance nginx_dev_1 {13     state MASTER14     interface eth115     mcast_src_ip 172.16.23.20316     virtual_router_id 10017     priority 10018     advert_int 519     authentication {20         auth_type PASS21         auth_pass 12345622     }23     virtual_ipaddress {24         172.16.23.22225     }26 }27 28 vrrp_instance nginx_dev_2 {29     state BACKUP30     interface eth131     mcast_src_ip 172.16.23.20332     virtual_router_id 10133     priority 20034     advert_int 535     authentication {36         auth_type PASS37         auth_pass 12345638     }39     virtual_ipaddress {40         172.16.23.22341     }42 }

　　172.16.23.204 配置如下

1 global_defs { 2     notification_email {  #指定Keepalived在发生事情的时候，发送邮件通知，每行一个地址 3         xxxx@aa.com 4     } 5     notification_email_from xxxx@aa.com   #指定发件人 6     smtp_server 192.168.8.208   #发送email的smtp地址 7     smtp_connect_timeout 30     #超时时间 8     router_id nginx_dev_1         #运行Keepalived的机器标识号，可以相同也可以不同 9     router_id nginx_dev_210 }11 12 vrrp_instance nginx_dev_2 {13     state MASTER14     interface eth115     mcast_src_ip 172.16.23.20416     virtual_router_id 10117     priority 10018     advert_int 519     authentication {20         auth_type PASS21         auth_pass 12345622     }23     virtual_ipaddress {24         172.16.23.22325     }26 }27 28 vrrp_instance nginx_dev_1 {29     state BACKUP30     interface eth131     mcast_src_ip 172.16.23.20432     virtual_router_id 10033     priority 20034     advert_int 535     authentication {36         auth_type PASS37         auth_pass 12345638     }39     virtual_ipaddress {40         172.16.23.22241     }42 }

　　就是这两份配置，基本是一样的，互为主备，里面一个比较重要修改的是virtual_router_id 

　　上面实现的效果是访问172.16.23.222 时先转到172.16.20.203主机上，访问172.16.23.223时转到172.16.20.204主机上。当其中一台主机宕机时，就会自动切换，切换到好的主机上，这个过程就几秒的时间。

　　这里是需要两个IP地址，需要客户端进行负载选择，这一步可以通过DNS进行分发处理。

1 root@debian-t6:/usr/local/nginx/html# ip addr 2 1: lo: 
     
       mtu 65536 qdisc noqueue state UNKNOWN group default  3     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 4     inet 127.0.0.1/8 scope host lo 5        valid_lft forever preferred_lft forever 6 2: eth1: 
      
        mtu 1500 qdisc pfifo_fast state UP group default  7     link/ether fc:aa:14:9d:d9:5a brd ff:ff:ff:ff:ff:ff 8     inet 172.16.23.204/24 brd 172.16.23.255 scope global eth1 9        valid_lft forever preferred_lft forever10     inet 172.16.23.223/32 scope global eth111        valid_lft forever preferred_lft forever12     inet 172.16.23.222/32 scope global eth113        valid_lft forever preferred_lft forever
      
     

　　日志在 /var/log/message

1 Jan 3 21:02:33 debian-t6 Keepalived_vrrp[31784]: VRRP_Instance(nginx_dev_1) Sending/queueing gratuitous ARPs on eth1 for 172.16.23.222 2 Jan 3 21:02:33 debian-t6 Keepalived_vrrp[31784]: Sending gratuitous ARP on eth1 for 172.16.23.222 3 Jan 3 21:02:33 debian-t6 Keepalived_vrrp[31784]: Sending gratuitous ARP on eth1 for 172.16.23.222 4 Jan 3 21:02:33 debian-t6 Keepalived_vrrp[31784]: Sending gratuitous ARP on eth1 for 172.16.23.222 5 Jan 3 21:02:33 debian-t6 Keepalived_vrrp[31784]: Sending gratuitous ARP on eth1 for 172.16.23.222 6 Jan 3 21:03:02 debian-t6 Keepalived_vrrp[31784]: VRRP_Instance(nginx_dev_1) sent 0 priority 7 Jan 3 21:03:02 debian-t6 Keepalived_healthcheckers[31783]: Stopped 8 Jan 3 21:03:02 debian-t6 Keepalived_vrrp[31784]: VRRP_Instance(nginx_dev_1) removing protocol VIPs. 9 Jan 3 21:03:04 debian-t6 Keepalived_vrrp[31784]: Stopped10 Jan 3 21:03:16 debian-t6 Keepalived_healthcheckers[31845]: Opening file '/etc/keepalived/keepalived.conf'.11 Jan 3 21:03:16 debian-t6 Keepalived_vrrp[31846]: Registering Kernel netlink reflector12 Jan 3 21:03:16 debian-t6 Keepalived_vrrp[31846]: Registering Kernel netlink command channel13 Jan 3 21:03:16 debian-t6 Keepalived_vrrp[31846]: Registering gratuitous ARP shared channel14 Jan 3 21:03:16 debian-t6 Keepalived_vrrp[31846]: Opening file '/etc/keepalived/keepalived.conf'.15 Jan 3 21:03:16 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) removing protocol VIPs.16 Jan 3 21:03:16 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_1) removing protocol VIPs.17 Jan 3 21:03:16 debian-t6 Keepalived_vrrp[31846]: Using LinkWatch kernel netlink reflector...18 Jan 3 21:03:16 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_1) Entering BACKUP STATE19 Jan 3 21:03:16 debian-t6 Keepalived_vrrp[31846]: VRRP sockpool: [ifindex(2), proto(112), unicast(0), fd(10,11)]20 Jan 3 21:03:18 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_1) forcing a new MASTER election21 Jan 3 21:03:20 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Transition to MASTER STATE22 Jan 3 21:03:20 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Received advert with higher priority 200, ours 10023 Jan 3 21:03:20 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Entering BACKUP STATE24 Jan 3 21:03:23 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Transition to MASTER STATE25 Jan 3 21:03:23 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_1) Transition to MASTER STATE26 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Entering MASTER STATE27 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) setting protocol VIPs.28 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.22329 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Sending/queueing gratuitous ARPs on eth1 for 172.16.23.22330 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.22331 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.22332 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.22333 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.22334 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_1) Entering MASTER STATE35 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_1) setting protocol VIPs.36 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.22237 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_1) Sending/queueing gratuitous ARPs on eth1 for 172.16.23.22238 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.22239 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.22240 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.22241 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.222

5. KeepAlived其他配置

　　下面讲解一些其他配置和高级应用

　　（1）email通知，这里就不做了，建议用第三方独立监控服务，如Nagios、Zabbix进行监控

　　（2）router_id 用户标识本节点名称，通常为hostname

　　（3）vrrp_instance 实例定义

　　（4）state 实例状态，只有MASTER 和 BACKUP两种状态，并且需要全部大写。抢占模式下其中MASTER为工作状态，BACKUP为备用状态。当MASTER所在服务器失效时，BACKUP所在服务器会自动把它的状态由BACKUP切换到MASTER状态。当失效的MASTER所在的服务恢复时，BACKUP从MASTER恢复到BACKUP状态

　　（5）interface 对望提供服务网卡接口，即VIP绑定的网卡接口。一些服务器都有两个以上的网卡接口，在选择接口时要正确选择

　　（6）mcast_src_ip 本机IP地址

　　（7）virtual_router_id 虚拟路由的ID号，每个节点必须设置一样。相同的VRID为一个组，将决定多播的MAC地址，不同的实例节点必须不一样

　　（8）priority 节点优先级，取值0~254， MASTER 要比BACKUP高

　　（9）advert_int MASTER 与 BACKUP 节点间同步检查的时间间隔，单位秒

　　（10）authentication 验证类型和验证密码。主要用PASS 密码模式 {auth_type PASS auth_pass 123456} 同一个vrrp实例MASTER与BACKUP使用相同的密码才能正常通信

　　（11）nopreempt 禁止抢占服务。默认情况，当MASTER服务挂掉之后，BACKUP自动升级为MASTER并接替其任务，当MASTER服务恢复后，升级为MASTER的BACKUP服务又自动降为BACKUP，把工作权交给原MASTER。当配置了nopreempt，MASTER从挂到到恢复，不再将服务抢占过来

　　（12）virtual_ipaddress 虚拟IP地址池，可以有多个IP，每个IP占一行，不需要指定子网掩码。注意这里的IP必须与我们设定的VIP保持一致 VRRP HA虚拟地址

　　（13）notify_master 表示当切换到master状态时，要执行的脚本

　　（14）notify_backup 表示当切换到backup状态时，要执行的脚本

　　（15）notify_fault 表示切换出现故障是要执行的脚本(这里也可以发送邮件什么的)

　　（16）track_script 执行监控的服务

　　（17）vrrp_script VRRP 脚本检测

6. KeepAlived 高级应用

　　172.16.23.203 

　　/etc/keepalived/keepalived.conf

1 global_defs { 2     router_id nginx_dev_1 3 } 4  5 #用于监控Nginx、Redis等应用是否在运行 6 vrrp_script chk_nginx_port { 7     script "/etc/keepalived/check.sh" #通过脚本检测，根据返回值进行判断 8     interval 2                        #脚本执行间隔 每2秒执行一次 9     weight -5                         #检测失败，优先级变更10     fall 2                            #连续检测2次失败才算失败11     rise 1                            #检测一次成功算成功，但不修改优先级12 }13 14 vrrp_instance nginx_dev_1 {15     state MASTER16     interface eth117     mcast_src_ip 172.16.23.20318     virtual_router_id 10019     priority 10020     advert_int 521     authentication {22         auth_type PASS23         auth_pass 12345624     }25     virtual_ipaddress {  #VRRP HA 虚拟地址，可以写多个26         172.16.23.22227     }28 29     notify_master "/etc/keepalived/run.sh master1"30     notify_backup "/etc/keepalived/run.sh backup1"31     notify_fault  "/etc/keepalived/run.sh fault1"32 33     track_script {34         chk_nginx_port     #执行对于脚本35     }36 }37 38 vrrp_instance nginx_dev_2 {39     state BACKUP40     interface eth141     mcast_src_ip 172.16.23.20342     virtual_router_id 10143     priority 20044     advert_int 545     authentication {46         auth_type PASS47         auth_pass 12345648     }49     virtual_ipaddress {50         172.16.23.223  #VRRP HA虚拟地址51     }52 53     notify_master "/etc/keepalived/run.sh master2"54     notify_backup "/etc/keepalived/run.sh backup2"55     notify_fault  "/etc/keepalived/run.sh fault2"56 }

　　/etc/keepalived/check.sh

1 #!/bin/bash  2 counter=$(ps -C nginx --no-heading | wc -l) 3 if [ "${counter}" = "0" ]; then 4     service nginx start 5     #sleep 2  # 这个在执行过程中有问题 6     counter=$(ps -C nginx --no-heading | wc -l) 7     if [ "${counter}" = "0" ]; then 8         service keepalived stop 9     fi10 fi

　　/etc/keepalived/run.sh

1 #!/bin/sh 2 echo $(date +%H:%M:%S) $1 >> /etc/keepalived/time.txt

　　172.16.23.204

　　/etc/keepalived/keepalived.conf

1 global_defs { 2     router_id nginx_dev_2         #运行Keepalived的机器标识号，可以相同也可以不同 3 } 4  5 #用于监控Nginx、Redis等应用是否在运行 6 vrrp_script chk_nginx_port { 7     script "/etc/keepalived/check.sh" 8     interval 2 9     weight -510     fall 211     rise 112 }13 14 vrrp_instance nginx_dev_2 {15     state MASTER16     interface eth117     mcast_src_ip 172.16.23.20418     virtual_router_id 10119     priority 10020     advert_int 521     authentication {22         auth_type PASS23         auth_pass 12345624     }25     virtual_ipaddress {26         172.16.23.22327     }28 29     notify_master "/etc/keepalived/run.sh master1"30     notify_backup "/etc/keepalived/run.sh backup1"31     notify_fault "/etc/keepalived/run.sh fault1"32 33     track_script {34         chk_nginx_port35     }36 }37 38 vrrp_instance nginx_dev_1 {39     state BACKUP40     interface eth141     mcast_src_ip 172.16.23.20442     virtual_router_id 10043     priority 20044     advert_int 545     authentication {46         auth_type PASS47         auth_pass 12345648     }49     virtual_ipaddress {50         172.16.23.22251     }52 53     notify_master "/etc/keepalived/run.sh master2"54     notify_backup "/etc/keepalived/run.sh backup2"55     notify_fault "/etc/keepalived/run.sh fault2"56 57     track_script {58         chk_nginx_port59     }60 }

　　/etc/keepalived/check.sh

1 #!/bin/bash 2 counter=$(ps -C nginx --no-heading | wc -l)3 if [ "${counter}" = "0" ]; then4     service nginx start5     counter=$(ps -C nginx --no-heading | wc -l)6     if [ "${counter}" = "0" ]; then7         service keepalived stop8     fi9 fi

　　/etc/keepalived/run.sh

1 #!/bin/sh 2 echo $(date +%H:%M:%S) $1 >> /etc/keepalived/time.txt

　　上面的那些sh文件要 chmod a+x *.sh ，在keepalived.conf 文件中最好是使用绝对路径

　　测试过程:

　　先启动203 的KeepAlived和Nginx

　　然后启动204的KeepAlived和Nginx

　　用浏览器访问 172.16.23.223 / 172.16.23.222 都是没有问题，基本被负载到两台主机上去了

　　场景一

　　　　然后204机器 service nginx stop && service nginx status 模拟异常退出并查看状态

　　　　过几秒后再查看204 service nginx status 发现nginx 自动重启了

　　场景二

　　　　修改check.sh 文件 把service nginx start 这一行注释掉

　　　　然后204 机器 service nginx stop && service nginx status 模拟异常退出并查看状态

　　　　过几秒后再查看204 service nginx status / service keepalived status 发现nginx关闭、对于的204上的KeepAlived也被关闭

　　　　这个时候，203服务器 tail -f time.txt 会出现一条master切换信息 表示VIP进行切换

　　　　浏览器访问 172.16.23.223/172.16.23.222 访问正常，不过都是指向同一个机器

　　　　取消check.sh 文件的注释

　　　　启动204的 KeepAlived service keepalived start 由于取消注释，对于的nginx也自动启动了

　　　　这时，203服务器 tail -f time.txt 会出现一条backup切换信息，表示203主机切换为Backup状态了。

　　　　这时浏览器访问以上两个IP，同样被负载的两台主机上