Keepalived 双机热备

使用 Keepalived 做双机热备非常简单，经常和 LVS 搭配来实现高可用负载平衡方案

1. Master / Slave

首先准备两台测试服务器和一个虚拟IP。

Server A: 192.168.1.10 (主服务器)

Server B: 192.168.1.20

Virtual IP: 192.168.1.100

测试服务: 在两台服务器上分别安装 Nginx，并修改默认的 index.html 文件，显示当前服务器 IP 以便识别。

1. 在两台服务器上分别安装 keepalived。

$ sudo apt-get install keepalived

2. 添加配置文件。

Server A

$ sudo vim /etc/keepalived/keepalived.conf

global_defs {

router_id LVS_DEVEL

}

vrrp_instance VI_1 {

state MASTER

interface eth0

virtual_router_id 51 # 保持主从服务器一致

priority 100         # 优先级 (主服务器较高)

advert_int 1         # 心跳广播间隔(秒)

authentication {

auth_type PASS

auth_pass 1111

}

virtual_ipaddress {

192.168.1.100 # 虚拟IP地址，可以多个。

}

}

Server B

$ sudo vim /etc/keepalived/keepalived.conf

global_defs {

router_id LVS_DEVEL

}

vrrp_instance VI_1 {

state BACKUP

interface eth0

virtual_router_id 51

priority 99

advert_int 1

authentication {

auth_type PASS

auth_pass 1111

}

virtual_ipaddress {

192.168.1.100

}

}

注意：备份服务器 Server B 配置中 state 要改成 BACKUP，同时调低 priority。

3. 启动两台服务器上的 keepalived 服务。

$ sudo service keepalived start

重启后可以使用 "ip a" 查看虚拟 IP 信息。

Server A

$ ip a

1: lo: <loopback,up,lower_up> mtu 16436 qdisc noqueue state UNKNOWN

link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00

inet 127.0.0.1/8 scope host lo

inet6 ::1/128 scope host

valid_lft forever preferred_lft forever

2: eth0: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state UNKNOWN qlen 1000

link/ether 00:0c:29:4c:e7:e7 brd ff:ff:ff:ff:ff:ff

inet 192.168.1.10/24 brd 192.168.1.255 scope global eth0

inet 192.168.1.100/24 scope global secondary eth0

inet6 fe80::20c:29ff:fe4c:e7e7/64 scope link

valid_lft forever preferred_lft forever

Server B

$ ip a

1: lo: <loopback,up,lower_up> mtu 16436 qdisc noqueue state UNKNOWN

link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00

inet 127.0.0.1/8 scope host lo

inet6 ::1/128 scope host

valid_lft forever preferred_lft forever

2: eth0: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state UNKNOWN qlen 1000

link/ether 00:0c:29:01:d8:16 brd ff:ff:ff:ff:ff:ff

inet 192.168.1.20/24 brd 192.168.1.255 scope global eth0

inet6 fe80::20c:29ff:fe01:d816/64 scope link

valid_lft forever preferred_lft forever

4. 在第三台机器上进行访问测试。

$ curl http://192.168.1.10

Welcome to nginx! 192.168.1.10

$ curl http://192.168.1.20

Welcome to nginx! 192.168.1.20

$ curl http://192.168.1.100

Welcome to nginx! 192.168.1.10

我们关掉主服务器 192.168.1.10，再访问 http://192.168.1.100 就会自动切换成备份服务器 (Server B: 192.168.1.20)。

$ curl http://192.168.1.100

Welcome to nginx! 192.168.1.20

同时 Server B 绑定了虚拟 IP。

Server B

$ ip a

1: lo: <loopback,up,lower_up> mtu 16436 qdisc noqueue state UNKNOWN

link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00

inet 127.0.0.1/8 scope host lo

inet6 ::1/128 scope host

valid_lft forever preferred_lft forever

2: eth0: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state UNKNOWN qlen 1000

link/ether 00:0c:29:01:d8:16 brd ff:ff:ff:ff:ff:ff

inet 192.168.1.20/24 brd 192.168.1.255 scope global eth0

inet 192.168.1.100/24 scope global secondary eth0

inet6 fe80::20c:29ff:fe01:d816/64 scope link

valid_lft forever preferred_lft forever

重新打开主服务器(Server A: 192.168.1.10)，访问恢复。

2. Master / Master

Master / Slave 方案中备份服务器(Server B)平时就是个摆设，有点浪费。我们完全可以用来跑其他服务，让两台主机形成相互热备。

Server A: 192.168.1.10, Virtual IP: 192.168.1.100

Server B: 192.168.1.20, Virtual IP: 192.168.1.200

修改配置文件。

Server A

global_defs {

router_id LVS_DEVEL

}

vrrp_instance VI_1 {

state MASTER

interface eth0

virtual_router_id 51

priority 100

advert_int 1

authentication {

auth_type PASS

auth_pass 1111

}

virtual_ipaddress {

192.168.1.100

}

}

vrrp_instance VI_2 {

state BACKUP

interface eth0

virtual_router_id 52

priority 99

advert_int 1

authentication {

auth_type PASS

auth_pass 1111

}

virtual_ipaddress {

192.168.1.200

}

}

Server B:

global_defs {

router_id LVS_DEVEL

}

vrrp_instance VI_1 {

state BACKUP

interface eth0

virtual_router_id 51

priority 99

advert_int 1

authentication {

auth_type PASS

auth_pass 1111

}

virtual_ipaddress {

192.168.1.100

}

}

vrrp_instance VI_2 {

state MASTER

interface eth0

virtual_router_id 52

priority 100

advert_int 1

authentication {

auth_type PASS

auth_pass 1111

}

virtual_ipaddress {

192.168.1.200

}

}

其实很简单，我们增加了一个新的配置 VI_2 (注意 virtual_router_id 不同)。不过这回用 Server B 做主服务器，如此 Server A、Server B 各自拥有主虚拟IP，同时备份对方的虚拟 IP。重启两台服务器的 keepalived 服务后，查看虚拟 IP 绑定信息。

Server A

$ ip a

1: lo: <loopback,up,lower_up> mtu 16436 qdisc noqueue state UNKNOWN

link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00

inet 127.0.0.1/8 scope host lo

inet6 ::1/128 scope host

valid_lft forever preferred_lft forever

2: eth0: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state UNKNOWN qlen 1000

link/ether 00:0c:29:4c:e7:e7 brd ff:ff:ff:ff:ff:ff

inet 192.168.1.10/24 brd 192.168.1.255 scope global eth0

inet 192.168.1.100/24 scope global secondary eth0

inet6 fe80::20c:29ff:fe4c:e7e7/64 scope link

valid_lft forever preferred_lft forever

Server B

$ ip a

1: lo: <loopback,up,lower_up> mtu 16436 qdisc noqueue state UNKNOWN

link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00

inet 127.0.0.1/8 scope host lo

inet6 ::1/128 scope host

valid_lft forever preferred_lft forever

2: eth0: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state UNKNOWN qlen 1000

link/ether 00:0c:29:01:d8:16 brd ff:ff:ff:ff:ff:ff

inet 192.168.1.20/24 brd 192.168.1.255 scope global eth0

inet 192.168.1.200/24 scope global secondary eth0

inet6 fe80::20c:29ff:fe01:d816/64 scope link

valid_lft forever preferred_lft forever

正常情况下，会使用各自的主服务器。

$ curl http://192.168.1.100

Welcome to nginx! 192.168.1.10

$ curl http://192.168.1.200

Welcome to nginx! 192.168.1.20

一旦任何一台服务器当机，另一台就会自动接管。我们停掉 192.168.1.20，看看访问 http://192.168.1.200 是不是切换到 192.168.1.10 上。

$ curl http://192.168.1.200

Welcome to nginx! 192.168.1.10

同时 Server A 绑定虚拟 IP 192.168.1.200。

$ ip a

1: lo: <loopback,up,lower_up> mtu 16436 qdisc noqueue state UNKNOWN

link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00

inet 127.0.0.1/8 scope host lo

inet6 ::1/128 scope host

valid_lft forever preferred_lft forever

2: eth0: <broadcast,multicast,up,lower_up> mtu 1500 qdisc pfifo_fast state UNKNOWN qlen 1000

link/ether 00:0c:29:4c:e7:e7 brd ff:ff:ff:ff:ff:ff

inet 192.168.1.10/24 brd 192.168.1.255 scope global eth0

inet 192.168.1.100/24 scope global secondary eth0

inet 192.168.1.200/24 scope global secondary eth0

inet6 fe80::20c:29ff:fe4c:e7e7/64 scope link

valid_lft forever preferred_lft forever

Server B 重启后，一切恢复正常。

这个方案可以是不同的服务，或者是同一服务的访问分流(配合 DNS 使用)。