一 环境介绍
1.操作系统
CentOS Linux release 7.2.1511 (Core)
2.服务
keepalived+nginx双主高可用负载均衡集群及LAMP应用
keepalived-1.2.13-7.el7.x86_64
nginx-1.10.2-1.el7.x86_64
httpd-2.4.6-45.el7.centos.x86_64
二 原理及拓扑图
1.vrrp协议
在现实的网络环境中,两台需要通信的主机大多数情况下并没有直接的物理连接。对于这样的情况,它们之间路由怎样选择?主机如何选定到达目的主机的下一跳路由,这个问题通常的解决方法有二种:
在主机上使用动态路由协议(RIP、OSPF等)
在主机上配置静态路由
很明显,在主机上配置动态路由是非常不切实际的,因为管理、维护成本以及是否支持等诸多问题。配置静态路由就变得十分流行,但路由器(或者说默认网关default gateway)却经常成为单点故障。VRRP的目的就是为了解决静态路由单点故障问题,VRRP通过一竞选(election)协议来动态的将路由任务交给LAN中虚拟路由器中的某台VRRP路由器。
2.nginx反代
nginx是以反向代理的方式进行负载均衡的。反向代理(Reverse Proxy)方式是指以代理服务器来接受Internet上的连接请求,然后将请求转发给内部网络上的服务器,并将从服务器上得到的结果返回给Internet上请求连接的客户端,此时代理服务器对外就表现为一个服务器。(为了理解反向代理,这里插播一条什么是正向代理:正向代理指的是,一个位于客户端和原始服务器之间的服务器,为了从原始服务器取得内容,客户端向代理发送一个请求并指定目标(原始服务器),然后代理向原始服务器转交请求并将获得的内容返回给客户端。)
3.拓扑图
三 配置
1.后端RS配置
1
2
|
[root@inode4 ~]
# yum install httpd -y
[root@inode5 ~]
# yum install httpd -y
|
2.Nginx反代配置
MASTER:
1
2
3
4
5
6
7
8
9
10
|
upstream websrvs {
server 172.18.67.11:80;
server 172.18.67.12:80;
server 127.0.0.1:80 backup;
}
server {
listen 80 ;
location / {
proxy_pass http:
//websrvs
;
}
|
BACKUP:
1
2
3
4
5
6
7
8
9
10
|
upstream websrvs {
server 172.18.67.11:80;
server 172.18.67.12:80;
server 127.0.0.1:80 backup;
}
server {
listen 80 ;
location / {
proxy_pass http:
//websrvs
;
}
|
3.keepalived高可用配置
MASTER:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
|
! Configuration File
for
keepalived
global_defs {
notification_email {
root@localhost
}
notification_email_from keepalived@localhost
smtp_server 127.0.0.1
smtp_connect_timeout 30
router_id node1
vrrp_mcast_group4 224.0.67.67
}
vrrp_script chk_down {
script
"[[ -f /etc/keepalived/down ]] && exit 1 || exit 0"
interval 1
weight -5
}
vrrp_script chk_nginx {
script
"killall -0 nginx && exit 0 || exit 1"
interval 1
weight -5
fall 2
rise 1
}
vrrp_instance myr {
state MASTER
interface eno16777736
virtual_router_id 167
priority 100
advert_int 1
authentication {
auth_type PASS
auth_pass 571f97b2
}
virtual_ipaddress {
172.18.67.33
/16
dev eno16777736
}
track_script {
chk_down
chk_nginx
}
notify_master
"/etc/keepalived/notify.sh master"
notify_backup
"/etc/keepalived/notify.sh backup"
notify_fault
"/etc/keepalived/notify.sh fault"
}
|
BACKUP:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
|
! Configuration File
for
keepalived
global_defs {
notification_email {
root@localhost
}
notification_email_from keepalived@localhost
smtp_server 127.0.0.1
smtp_connect_timeout 30
router_id node1
vrrp_mcast_group4 224.0.67.67
}
vrrp_script chk_down {
script
"[[ -f /etc/keepalived/down ]] && exit 1 || exit 0"
interval 1
weight -5
}
vrrp_script chk_nginx {
script
"killall -0 nginx && exit 0 || exit 1"
interval 1
weight -5
fall 2
rise 1
}
vrrp_instance myr {
state BACKUP
interface eno16777736
virtual_router_id 167
priority 95
advert_int 1
authentication {
auth_type PASS
auth_pass 571f97b2
}
virtual_ipaddress {
172.18.67.33
/16
dev eno16777736
}
track_script {
chk_down
chk_nginx
}
notify_master
"/etc/keepalived/notify.sh master"
notify_backup
"/etc/keepalived/notify.sh backup"
notify_fault
"/etc/keepalived/notify.sh fault"
}
|
4.通知脚本示例
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
|
[root@inode2 nginx]
# vim notify.sh
#!/bin/bash
#
contact=
'root@localhost'
notify() {
mailsubject=
"$(hostname) to be $1, vip floating"
mailbody=
"$(date +'%F %T'): vrrp transition, $(hostname) changed to be $1"
echo
"$mailbody"
| mail -s
"$mailsubject"
$contact
}
case
$1
in
master)
notify master
;;
backup)
notify backup
;;
fault)
notify fault
;;
*)
echo
"Usage: $(basename $0) {master|backup|fault}"
exit
1
;;
esac
|
节点二同样配置
四 启动服务并测试
1.启动后端web服务器
1
2
|
[root@inode4 ~]
# systemctl start httpd
[root@inode5 ~]
# systemctl start httpd
|
为了测试显示效果明显一点,自定义一个访问页面
1
2
|
[root@inode4 ~]
# echo "RS1:172.18.67.11" > /var/www/html/index.html
[root@inode5 ~]
# echo "RS2:172.18.67.12" > /var/www/html/index.html
|
2.测试
MASTER:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
|
[root@inode2 ~]
# systemctl start keepalived
[root@inode2 ~]
# systemctl status -l keepalived
● keepalived.service - LVS and VRRP High Availability Monitor
Loaded: loaded (
/usr/lib/systemd/system/keepalived
.service; disabled; vendor preset: disabled)
Active: active (running) since Mon 2017-05-15 15:45:20 CST; 3s ago
Process: 20971 ExecStart=
/usr/sbin/keepalived
$KEEPALIVED_OPTIONS (code=exited, status=0
/SUCCESS
)
Main PID: 20972 (keepalived)
CGroup:
/system
.slice
/keepalived
.service
├─20972
/usr/sbin/keepalived
-D
├─20973
/usr/sbin/keepalived
-D
└─20974
/usr/sbin/keepalived
-D
May 15 15:45:20 inode2 Keepalived_healthcheckers[20973]: Opening
file
'/etc/keepalived/keepalived.conf'
.
May 15 15:45:20 inode2 Keepalived_healthcheckers[20973]: Configuration is using : 7521 Bytes
May 15 15:45:20 inode2 Keepalived_healthcheckers[20973]: Using LinkWatch kernel netlink reflector...
May 15 15:45:20 inode2 Keepalived_vrrp[20974]: VRRP_Script(chk_nginx) succeeded
May 15 15:45:21 inode2 Keepalived_vrrp[20974]: VRRP_Instance(myr) Transition to MASTER STATE
May 15 15:45:22 inode2 Keepalived_vrrp[20974]: VRRP_Instance(myr) Entering MASTER STATE
May 15 15:45:22 inode2 Keepalived_vrrp[20974]: VRRP_Instance(myr) setting protocol VIPs.
May 15 15:45:22 inode2 Keepalived_vrrp[20974]: VRRP_Instance(myr) Sending gratuitous ARPs on eno16777736
for
172.18.67.33
May 15 15:45:22 inode2 Keepalived_vrrp[20974]: Opening script
file
/etc/keepalived/notify
.sh
May 15 15:45:22 inode2 Keepalived_healthcheckers[20973]: Netlink reflector reports IP 172.18.67.33 added
[root@inode2 ~]
# ip a l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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
valid_lft forever preferred_lft forever
inet6 ::1
/128
scope host
valid_lft forever preferred_lft forever
2: eno16777736: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link
/ether
00:0c:29:8b:08:6f brd ff:ff:ff:ff:ff:ff
inet 172.18.67.13
/16
brd 172.18.255.255 scope global eno16777736
valid_lft forever preferred_lft forever
inet 172.18.67.33
/16
scope global secondary eno16777736
valid_lft forever preferred_lft forever
inet6 fe80::20c:29ff:fe8b:86f
/64
scope link tentative dadfailed
valid_lft forever preferred_lft forever
|
主节点启动,Entering MASTER STATE,此时我们在客户端进行测试访问
1
2
3
4
5
|
[root@inode1 ~]
# for i in {1..4};do curl http://172.18.67.33;done
RS1:172.18.67.11
RS2:172.18.67.12
RS1:172.18.67.11
RS2:172.18.67.12
|
访问正常,接下来我们启动备用节点的服务器
BACKUP:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
|
[root@inode3 keepalived]
# systemctl start keepalived
[root@inode3 keepalived]
# systemctl status -l keepalived
● keepalived.service - LVS and VRRP High Availability Monitor
Loaded: loaded (
/usr/lib/systemd/system/keepalived
.service; disabled; vendor preset: disabled)
Active: active (running) since Mon 2017-05-15 15:46:51 CST; 3s ago
Process: 24329 ExecStart=
/usr/sbin/keepalived
$KEEPALIVED_OPTIONS (code=exited, status=0
/SUCCESS
)
Main PID: 24330 (keepalived)
CGroup:
/system
.slice
/keepalived
.service
├─24330
/usr/sbin/keepalived
-D
├─24331
/usr/sbin/keepalived
-D
└─24332
/usr/sbin/keepalived
-D
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: Registering Kernel netlink
command
channel
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: Registering gratuitous ARP shared channel
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: Opening
file
'/etc/keepalived/keepalived.conf'
.
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: Configuration is using : 66427 Bytes
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: Using LinkWatch kernel netlink reflector...
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: VRRP_Instance(myr) Entering BACKUP STATE
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: Opening script
file
/etc/keepalived/notify
.sh
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: VRRP sockpool: [ifindex(2), proto(112), unicast(0), fd(10,11)]
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: VRRP_Script(chk_down) succeeded
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: VRRP_Script(chk_nginx) succeeded
[root@inode3 keepalived]
# ip a l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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
valid_lft forever preferred_lft forever
inet6 ::1
/128
scope host
valid_lft forever preferred_lft forever
2: eno16777736: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link
/ether
00:0c:29:78:24:c3 brd ff:ff:ff:ff:ff:ff
inet 172.18.67.14
/16
brd 172.18.255.255 scope global eno16777736
valid_lft forever preferred_lft forever
inet6 fe80::20c:29ff:fe78:24c3
/64
scope link tentative dadfailed
valid_lft forever preferred_lft forever
|
此时,我们可以看到备用节点服务器启动后进入了BACKUP状态,Entering BACKUP STATE。接下来我们测试主节点宕机的情形下,我们的服务是否还可用
1
|
[root@inode2 ~]
# systemctl stop keepalived
|
主节点宕机后我们查看备用节点的状态
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
|
[root@inode3 keepalived]
# systemctl status -l keepalived
● keepalived.service - LVS and VRRP High Availability Monitor
Loaded: loaded (
/usr/lib/systemd/system/keepalived
.service; disabled; vendor preset: disabled)
Active: active (running) since Mon 2017-05-15 15:46:51 CST; 2min 19s ago
Process: 24329 ExecStart=
/usr/sbin/keepalived
$KEEPALIVED_OPTIONS (code=exited, status=0
/SUCCESS
)
Main PID: 24330 (keepalived)
CGroup:
/system
.slice
/keepalived
.service
├─24330
/usr/sbin/keepalived
-D
├─24331
/usr/sbin/keepalived
-D
└─24332
/usr/sbin/keepalived
-D
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: VRRP sockpool: [ifindex(2), proto(112), unicast(0), fd(10,11)]
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: VRRP_Script(chk_down) succeeded
May 15 15:46:51 inode3 Keepalived_vrrp[24332]: VRRP_Script(chk_nginx) succeeded
May 15 15:48:35 inode3 Keepalived_vrrp[24332]: VRRP_Instance(myr) Transition to MASTER STATE
May 15 15:48:36 inode3 Keepalived_vrrp[24332]: VRRP_Instance(myr) Entering MASTER STATE
May 15 15:48:36 inode3 Keepalived_vrrp[24332]: VRRP_Instance(myr) setting protocol VIPs.
May 15 15:48:36 inode3 Keepalived_vrrp[24332]: VRRP_Instance(myr) Sending gratuitous ARPs on eno16777736
for
172.18.67.33
May 15 15:48:36 inode3 Keepalived_vrrp[24332]: Opening script
file
/etc/keepalived/notify
.sh
May 15 15:48:36 inode3 Keepalived_healthcheckers[24331]: Netlink reflector reports IP 172.18.67.33 added
May 15 15:48:41 inode3 Keepalived_vrrp[24332]: VRRP_Instance(myr) Sending gratuitous ARPs on eno16777736
for
172.18.67.33
[root@inode3 keepalived]
# ip a l
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 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
valid_lft forever preferred_lft forever
inet6 ::1
/128
scope host
valid_lft forever preferred_lft forever
2: eno16777736: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link
/ether
00:0c:29:78:24:c3 brd ff:ff:ff:ff:ff:ff
inet 172.18.67.14
/16
brd 172.18.255.255 scope global eno16777736
valid_lft forever preferred_lft forever
inet 172.18.67.33
/16
scope global secondary eno16777736
valid_lft forever preferred_lft forever
inet6 fe80::20c:29ff:fe78:24c3
/64
scope link tentative dadfailed
valid_lft forever preferred_lft forever
|
我们发现备用节点由备用状态进入了主状态,并且IP地址也成功绑定至备用节点下。再次进行测试访问
1
2
3
4
5
|
[root@inode1 ~]
# for i in {1..4};do curl http://172.18.67.33;done
RS1:172.18.67.11
RS2:172.18.67.12
RS1:172.18.67.11
RS2:172.18.67.12
|
测试一台web服务器宕机
1
2
3
4
5
6
|
[root@inode4 ~]
# systemctl stop httpd
[root@inode1 ~]
# for i in {1..4};do curl http://172.18.67.33;done
RS2:172.18.67.12
RS2:172.18.67.12
RS2:172.18.67.12
RS2:172.18.67.12
|
在实际生产环境中后端两台web服务器的内容应该一样的,在这里我们可认为客户端已成功访问到服务器,因此我们可认为这样的架构体现了高可用负载均衡。