物联网架构成长之路-双机热备Keepalived了解

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 make
2 make install
3 mkdir /etc/keepalived
4 cp ./etc/keepalived/keepalived.conf /etc/keepalived/keepalived.conf
5 cp ./sbin/keepalived /usr/sbin/
6 vim /etc/init.d/keepalived
7 chmod a+x /etc/init.d/keepalived
8 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         [email protected]
 4 　　}
 5     notification_email_from [email protected] #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 MASTER
13     interface eth1
14     #mcast_src_ip 172.16.23.203
15     virtual_router_id 100
16     priority 200
17     advert_int 5
18     authentication {
19         auth_type PASS
20         auth_pass 123456
21     }
22     virtual_ipaddress {
23         172.16.23.222
24     }
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         [email protected]
 4     }
 5     notification_email_from [email protected]   #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_2
10 }
11 
12 vrrp_instance nginx_dev_1 {
13     state MASTER
14     interface eth1
15     mcast_src_ip 172.16.23.203
16     virtual_router_id 100
17     priority 100
18     advert_int 5
19     authentication {
20         auth_type PASS
21         auth_pass 123456
22     }
23     virtual_ipaddress {
24         172.16.23.222
25     }
26 }
27 
28 vrrp_instance nginx_dev_2 {
29     state BACKUP
30     interface eth1
31     mcast_src_ip 172.16.23.203
32     virtual_router_id 101
33     priority 200
34     advert_int 5
35     authentication {
36         auth_type PASS
37         auth_pass 123456
38     }
39     virtual_ipaddress {
40         172.16.23.223
41     }
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         [email protected]
 4     }
 5     notification_email_from [email protected]   #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_2
10 }
11 
12 vrrp_instance nginx_dev_2 {
13     state MASTER
14     interface eth1
15     mcast_src_ip 172.16.23.204
16     virtual_router_id 101
17     priority 100
18     advert_int 5
19     authentication {
20         auth_type PASS
21         auth_pass 123456
22     }
23     virtual_ipaddress {
24         172.16.23.223
25     }
26 }
27 
28 vrrp_instance nginx_dev_1 {
29     state BACKUP
30     interface eth1
31     mcast_src_ip 172.16.23.204
32     virtual_router_id 100
33     priority 200
34     advert_int 5
35     authentication {
36         auth_type PASS
37         auth_pass 123456
38     }
39     virtual_ipaddress {
40         172.16.23.222
41     }
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: <LOOPBACK,UP,LOWER_UP> 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: <BROADCAST,MULTICAST,UP,LOWER_UP> 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 forever
10     inet 172.16.23.223/32 scope global eth1
11        valid_lft forever preferred_lft forever
12     inet 172.16.23.222/32 scope global eth1
13        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]: Stopped
10 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 reflector
12 Jan 3 21:03:16 debian-t6 Keepalived_vrrp[31846]: Registering Kernel netlink command channel
13 Jan 3 21:03:16 debian-t6 Keepalived_vrrp[31846]: Registering gratuitous ARP shared channel
14 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 STATE
19 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 election
21 Jan 3 21:03:20 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Transition to MASTER STATE
22 Jan 3 21:03:20 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Received advert with higher priority 200, ours 100
23 Jan 3 21:03:20 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Entering BACKUP STATE
24 Jan 3 21:03:23 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Transition to MASTER STATE
25 Jan 3 21:03:23 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_1) Transition to MASTER STATE
26 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Entering MASTER STATE
27 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 172.16.23.223
29 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_2) Sending/queueing gratuitous ARPs on eth1 for 172.16.23.223
30 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.223
31 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.223
32 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.223
33 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.223
34 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_1) Entering MASTER STATE
35 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 172.16.23.222
37 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: VRRP_Instance(nginx_dev_1) Sending/queueing gratuitous ARPs on eth1 for 172.16.23.222
38 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.222
39 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.222
40 Jan 3 21:03:28 debian-t6 Keepalived_vrrp[31846]: Sending gratuitous ARP on eth1 for 172.16.23.222
41 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 MASTER
16     interface eth1
17     mcast_src_ip 172.16.23.203
18     virtual_router_id 100
19     priority 100
20     advert_int 5
21     authentication {
22         auth_type PASS
23         auth_pass 123456
24     }
25     virtual_ipaddress {  #VRRP HA 虚拟地址，可以写多个
26         172.16.23.222
27     }
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 BACKUP
40     interface eth1
41     mcast_src_ip 172.16.23.203
42     virtual_router_id 101
43     priority 200
44     advert_int 5
45     authentication {
46         auth_type PASS
47         auth_pass 123456
48     }
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     fi
10 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 -5
10     fall 2
11     rise 1
12 }
13 
14 vrrp_instance nginx_dev_2 {
15     state MASTER
16     interface eth1
17     mcast_src_ip 172.16.23.204
18     virtual_router_id 101
19     priority 100
20     advert_int 5
21     authentication {
22         auth_type PASS
23         auth_pass 123456
24     }
25     virtual_ipaddress {
26         172.16.23.223
27     }
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
35     }
36 }
37 
38 vrrp_instance nginx_dev_1 {
39     state BACKUP
40     interface eth1
41     mcast_src_ip 172.16.23.204
42     virtual_router_id 100
43     priority 200
44     advert_int 5
45     authentication {
46         auth_type PASS
47         auth_pass 123456
48     }
49     virtual_ipaddress {
50         172.16.23.222
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 
57     track_script {
58         chk_nginx_port
59     }
60 }

　　/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     counter=$(ps -C nginx --no-heading | wc -l)
6     if [ "${counter}" = "0" ]; then
7         service keepalived stop
8     fi
9 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