我们知道Keepalived原生设计目的是为了高可用LVS集群的，但Keepalived除了可以高可用LVS服务之外，还可以基于vrrp_script和track_script高可用其它服务，如Nginx等。本篇主要演示如何使用Keepalived高可用Nginx服务（双实例），关于vrrp_script、track_script的更多介绍可以见上一篇博客《Keepalived学习总结》。

实验要求 ==> 实现Keepalived基于vrrp_script、track_script高可用Nginx服务，要求配置为双实例，两个vrrp实例的VIP分别为192.168.10.77、192.168.10.78，可以分别将其配置在网卡别名上。

实验环境 ==> CentOS 7.x

实验目的 ==> 实现Keepalived双实例高可用Nginx服务

实验主机 ==> 192.168.10.6（主机名：node1）、192.168.10.8（主机名：node2）

实验前提 ==> 高可用对之间时间同步（可通过周期性任务来实现）

实验操作如下。

1、编辑Keepalived配置文件

2、在两个节点上分别使用yum安装Nginx，并启动Web服务

(1) 在node1（192.168.10.6）上

[root@node1 ~]# yum -y install nginx
[root@node1 ~]# systemctl start nginx.service
[root@node1 ~]# ss -tnl | grep :80
LISTEN     0      128          *:80                       *:*
LISTEN     0      128         :::80                      :::*

(2) 在node2（192.168.10.8）上

[root@node2 ~]# yum -y install nginx
[root@node2 ~]# systemctl start nginx.service
[root@node2 ~]# ss -tnl | grep :80
LISTEN     0      128          *:80                       *:*

3、在两个节点上分别启动Keepalived服务

(1) 在node1（192.168.10.6）上

[root@node1 ~]# systemctl start keepalived
[root@node1 ~]# ifconfig
ens33: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.6  netmask 255.255.255.0  broadcast 192.168.10.255ether 00:0c:29:f7:b3:4e  txqueuelen 1000  (Ethernet)RX packets 46064  bytes 21104876 (20.1 MiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 43548  bytes 3735943 (3.5 MiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0ens33:0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.77  netmask 255.255.255.255  broadcast 0.0.0.0ether 00:0c:29:f7:b3:4e  txqueuelen 1000  (Ethernet)lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536inet 127.0.0.1  netmask 255.0.0.0inet6 ::1  prefixlen 128  scopeid 0x10<host>loop  txqueuelen 0  (Local Loopback)RX packets 195  bytes 12522 (12.2 KiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 195  bytes 12522 (12.2 KiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

(2) 在node2（192.168.10.8）上

[root@node2 ~]# systemctl start keepalived
[root@node2 ~]# ifconfig
ens34: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.8  netmask 255.255.255.0  broadcast 192.168.10.255ether 00:0c:29:ef:52:87  txqueuelen 1000  (Ethernet)RX packets 60646  bytes 17433464 (16.6 MiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 29539  bytes 2636829 (2.5 MiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0ens34:1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.78  netmask 255.255.255.255  broadcast 0.0.0.0ether 00:0c:29:ef:52:87  txqueuelen 1000  (Ethernet)lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536inet 127.0.0.1  netmask 255.0.0.0inet6 ::1  prefixlen 128  scopeid 0x10<host>loop  txqueuelen 0  (Local Loopback)RX packets 319  bytes 20447 (19.9 KiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 319  bytes 20447 (19.9 KiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

可以发现，因为在第一个vrrp实例（VIP为192.168.10.77）中，node1节点为主节点，因此该VIP配置在node1上；而在第二个vrrp实例（VIP为192.168.10.78）中，node2节点为主节点，因此该VIP配置在node2上。

4、人为制造故障进行测试

(1) 在node2上杀死nginx进程，查看第二个vrrp实例的VIP（192.168.10.78）是否会漂移至node1。

[root@node2 ~]# killall nginx

#查看日志

[root@node2 ~]# tail /var/log/messages
...（其他省略）...
Aug  8 16:29:28 node2 Keepalived_vrrp[6832]: VRRP_Script(chk_nginx) failed
Aug  8 16:29:30 node2 Keepalived_vrrp[6832]: VRRP_Instance(VI_2) Received higher prio advert
Aug  8 16:29:30 node2 Keepalived_vrrp[6832]: VRRP_Instance(VI_2) Entering BACKUP STATE
Aug  8 16:29:30 node2 Keepalived_vrrp[6832]: VRRP_Instance(VI_2) removing protocol VIPs.
Aug  8 16:29:30 node2 Keepalived_healthcheckers[6831]: Netlink reflector reports IP 192.168.10.78 removed

#查看node2上的IP地址

[root@node2 ~]# ifconfig
ens34: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.8  netmask 255.255.255.0  broadcast 192.168.10.255ether 00:0c:29:ef:52:87  txqueuelen 1000  (Ethernet)RX packets 61483  bytes 17487867 (16.6 MiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 29994  bytes 2672261 (2.5 MiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536inet 127.0.0.1  netmask 255.0.0.0inet6 ::1  prefixlen 128  scopeid 0x10<host>loop  txqueuelen 0  (Local Loopback)RX packets 323  bytes 20647 (20.1 KiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 323  bytes 20647 (20.1 KiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

可以发现第二个vrrp实例的VIP（192.168.10.78）已经从node2上移除。

#查看node1上的IP地址

[root@node1 ~]# ifconfig
ens33: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.6  netmask 255.255.255.0  broadcast 192.168.10.255ether 00:0c:29:f7:b3:4e  txqueuelen 1000  (Ethernet)RX packets 46700  bytes 21146944 (20.1 MiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 44417  bytes 3794925 (3.6 MiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0ens33:0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.77  netmask 255.255.255.255  broadcast 0.0.0.0ether 00:0c:29:f7:b3:4e  txqueuelen 1000  (Ethernet)ens33:1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.78  netmask 255.255.255.255  broadcast 0.0.0.0ether 00:0c:29:f7:b3:4e  txqueuelen 1000  (Ethernet)lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536inet 127.0.0.1  netmask 255.0.0.0inet6 ::1  prefixlen 128  scopeid 0x10<host>loop  txqueuelen 0  (Local Loopback)RX packets 199  bytes 12722 (12.4 KiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 199  bytes 12722 (12.4 KiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

此时node1不仅是第一个vrrp实例（VIP为192.168.10.77）的主节点，还是第二个vrrp实例（VIP为192.168.10.78）的主节点。

(2) 在node1上杀死nginx进程，在node2上重新启动nginx进程，查看第一个vrrp实例的VIP（192.168.10.77）和第二个vrrp实例的VIP（192.168.10.78）是否会漂移至node2。

#先在node2上重启nginx进程，此时第二个vrrp实例的VIP（192.168.10.78）应该会重新漂移至node2上

[root@node2 ~]# systemctl restart nginx.service
[root@node2 ~]# ss -tnl | grep :80
LISTEN     0      128          *:80                       *:*
[root@node2 ~]# ifconfig
ens34: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.8  netmask 255.255.255.0  broadcast 192.168.10.255ether 00:0c:29:ef:52:87  txqueuelen 1000  (Ethernet)RX packets 62438  bytes 17551143 (16.7 MiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 30145  bytes 2688809 (2.5 MiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0ens34:1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.78  netmask 255.255.255.255  broadcast 0.0.0.0ether 00:0c:29:ef:52:87  txqueuelen 1000  (Ethernet)lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536inet 127.0.0.1  netmask 255.0.0.0inet6 ::1  prefixlen 128  scopeid 0x10<host>loop  txqueuelen 0  (Local Loopback)RX packets 331  bytes 21047 (20.5 KiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 331  bytes 21047 (20.5 KiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

果然在意料之中！

#在node1上杀死nginx进程，查看第一个vrrp实例的VIP（192.168.10.77）是否会重新漂移至node2上

[root@node1 ~]# killall nginx        # 杀死nginx进程
[root@node1 ~]# tail /var/log/messages    # 查看日志
Aug  8 16:36:41 node1 Keepalived_vrrp[7026]: VRRP_Instance(VI_2) Sending gratuitous ARPs on ens33 for 192.168.10.78
Aug  8 16:37:10 node1 Keepalived_vrrp[7026]: VRRP_Instance(VI_2) Received higher prio advert
Aug  8 16:37:10 node1 Keepalived_vrrp[7026]: VRRP_Instance(VI_2) Entering BACKUP STATE
Aug  8 16:37:10 node1 Keepalived_vrrp[7026]: VRRP_Instance(VI_2) removing protocol VIPs.
Aug  8 16:37:10 node1 Keepalived_healthcheckers[7025]: Netlink reflector reports IP 192.168.10.78 removed
Aug  8 16:38:44 node1 Keepalived_vrrp[7026]: VRRP_Script(chk_nginx) failed
Aug  8 16:38:45 node1 Keepalived_vrrp[7026]: VRRP_Instance(VI_1) Received higher prio advert
Aug  8 16:38:45 node1 Keepalived_vrrp[7026]: VRRP_Instance(VI_1) Entering BACKUP STATE
Aug  8 16:38:45 node1 Keepalived_vrrp[7026]: VRRP_Instance(VI_1) removing protocol VIPs.
Aug  8 16:38:45 node1 Keepalived_healthcheckers[7025]: Netlink reflector reports IP 192.168.10.77 removed
[root@node1 ~]# ifconfig     # 查看IP地址
ens33: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.6  netmask 255.255.255.0  broadcast 192.168.10.255ether 00:0c:29:f7:b3:4e  txqueuelen 1000  (Ethernet)RX packets 47288  bytes 21191119 (20.2 MiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 45193  bytes 3854623 (3.6 MiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536inet 127.0.0.1  netmask 255.0.0.0inet6 ::1  prefixlen 128  scopeid 0x10<host>loop  txqueuelen 0  (Local Loopback)RX packets 215  bytes 13522 (13.2 KiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 215  bytes 13522 (13.2 KiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

两个vrrp实例的VIP都不见了。

#重新查看node2上的IP地址

[root@node2 ~]# ifconfig
ens34: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.8  netmask 255.255.255.0  broadcast 192.168.10.255ether 00:0c:29:ef:52:87  txqueuelen 1000  (Ethernet)RX packets 62612  bytes 17562546 (16.7 MiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 30521  bytes 2712799 (2.5 MiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0ens34:0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.77  netmask 255.255.255.255  broadcast 0.0.0.0ether 00:0c:29:ef:52:87  txqueuelen 1000  (Ethernet)ens34:1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500inet 192.168.10.78  netmask 255.255.255.255  broadcast 0.0.0.0ether 00:0c:29:ef:52:87  txqueuelen 1000  (Ethernet)lo: flags=73<UP,LOOPBACK,RUNNING>  mtu 65536inet 127.0.0.1  netmask 255.0.0.0inet6 ::1  prefixlen 128  scopeid 0x10<host>loop  txqueuelen 0  (Local Loopback)RX packets 335  bytes 21247 (20.7 KiB)RX errors 0  dropped 0  overruns 0  frame 0TX packets 335  bytes 21247 (20.7 KiB)TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

两个vrrp实例的VIP都漂移到node2去了。

当然，本配置示例中自定义的vrrp_script过于简单，此处提供一个vrrp_script监控脚本，仅供参考。

[root@node1 ~]# vim /etc/keepalived/vrrp_script.sh
#!/bin/bash
#
server=nginxrestart_server() {if ! systemctl restart $server &> /dev/null; thensleep 1return 0elseexit 0fi
}if ! killall -0 $server &> /dev/null; thenrestart_serverrestart_serverrestart_server[ $? -ne 0 ] && exit 1 || exit 0
fi

在keepalived.conf配置文件中需要指明脚本路径。

[root@node1 ~]# vim /etc/keepalived/keepalived.conf
...（其他省略）...
vrrp_script chk_nginx {script "/etc/keepalived/vrrp_script.sh"interval 2weight -5
}
...（其他省略）...

实验完成！