如何快速构建高可用集群(Keepalived+Haproxy+Nginx)

  1. 云栖社区>
  2. 博客>
  3. 正文

如何快速构建高可用集群(Keepalived+Haproxy+Nginx)

余二五 2017-11-15 16:40:00 浏览1876
展开阅读全文

组件及实现的功能

  • Keepalived:实现对Haproxy服务的高可用,并采用双主模型配置;

  • Haproxy:实现对Nginx的负载均衡和读写分离;

  • Nginx:实现对HTTP请求的高速处理;


架构设计图

wKiom1Ni8taDxbQnAANjCL8rEcc897.jpg


重点概念

vrrp_script中节点权重改变算法

vrrp_script 里的script返回值为0时认为检测成功,其它值都会当成检测失败;

weight 为正时,脚本检测成功时此weight会加到priority上,检测失败时不加;

主失败:

主 priority < 从 priority + weight 时会切换。

主成功:

主 priority + weight > 从 priority + weight 时,主依然为主

weight 为负时,脚本检测成功时此weight不影响priority,检测失败时priority – abs(weight)

主失败:

主 priority – abs(weight) < 从priority 时会切换主从

主成功:

主 priority > 从priority 主依然为主

具体解释详见博文“Keepalived双主模型中vrrp_script中权重改变故障排查


部署配置

Keepalived部署

配置

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
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
yum -y install keepalived # 两节点都需部署
# 172.16.25.109
# vi /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
   notification_email {
         root@localhost
   }
   notification_email_from admin@lnmmp.com
   smtp_connect_timeout 3
   smtp_server 127.0.0.1
   router_id LVS_DEVEL
}
vrrp_script chk_maintaince_down {
   script "[[ -f /etc/keepalived/down ]] && exit 1 || exit 0"
   interval 1
   weight 2
}
vrrp_script chk_haproxy {
    script "killall -0 haproxy"
    interval 1
    weight 2
}
vrrp_instance VI_1 {
    interface eth0
    state MASTER
    priority 100
    virtual_router_id 125
    garp_master_delay 1
    authentication {
        auth_type PASS
        auth_pass 1e3459f77aba4ded
    }
    track_interface {
       eth0
    }
    virtual_ipaddress {
        172.16.25.10/16 dev eth0 label eth0:0
    }
    track_script {
        chk_haproxy
    }
    notify_master "/etc/keepalived/notify.sh master 172.16.25.10"
    notify_backup "/etc/keepalived/notify.sh backup 172.16.25.10"
    notify_fault "/etc/keepalived/notify.sh fault 172.16.25.10"
}
vrrp_instance VI_2 {
    interface eth0
    state BACKUP
    priority 99
    virtual_router_id 126
    garp_master_delay 1
    authentication {
        auth_type PASS
        auth_pass 7615c4b7f518cede
    }
    track_interface {
       eth0
    }
    virtual_ipaddress {
        172.16.25.11/16 dev eth0 label eth0:1
    }
    track_script {
        chk_haproxy
chk_maintaince_down
    }
    notify_master "/etc/keepalived/notify.sh master 172.16.25.11"
    notify_backup "/etc/keepalived/notify.sh backup 172.16.25.11"
    notify_fault "/etc/keepalived/notify.sh fault 172.16.25.11"
}
# 172.16.25.110
# vi /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
   notification_email {
         root@localhost
   }
   notification_email_from admin@lnmmp.com
   smtp_connect_timeout 3
   smtp_server 127.0.0.1
   router_id LVS_DEVEL
}
vrrp_script chk_maintaince_down {
   script "[[ -f /etc/keepalived/down ]] && exit 1 || exit 0"
   interval 1
   weight 2
}
vrrp_script chk_haproxy {
    script "killall -0 haproxy"
    interval 1
    weight 2
}
vrrp_instance VI_1 {
    interface eth0
    state BACKUP
    priority 99
    virtual_router_id 125
    garp_master_delay 1
    authentication {
        auth_type PASS
        auth_pass 1e3459f77aba4ded
    }
    track_interface {
       eth0
    }
    virtual_ipaddress {
        172.16.25.10/16 dev eth0 label eth0:0
    }
    track_script {
        chk_haproxy
chk_maintaince_down
    }
    notify_master "/etc/keepalived/notify.sh master 172.16.25.10"
    notify_backup "/etc/keepalived/notify.sh backup 172.16.25.10"
    notify_fault "/etc/keepalived/notify.sh fault 172.16.25.10"
}
vrrp_instance VI_2 {
    interface eth0
    state MASTER
    priority 100
    virtual_router_id 126
    garp_master_delay 1
    authentication {
        auth_type PASS
        auth_pass 7615c4b7f518cede
    }
    track_interface {
       eth0
    }
    virtual_ipaddress {
        172.16.25.11/16 dev eth0 label eth0:1
    }
    track_script {
        chk_haproxy
    }
    notify_master "/etc/keepalived/notify.sh master 172.16.25.11"
    notify_backup "/etc/keepalived/notify.sh backup 172.16.25.11"
    notify_fault "/etc/keepalived/notify.sh fault 172.16.25.11"
}
# vi /etc/keepalived/notify.sh
#!/bin/bash
# Author: Jason.Yu <admin@lnmmp.com>
# description: An example of notify script
#
contact='root@localhost'
notify() {
    mailsubject="`hostname` to be $1: $2 floating"
    mailbody="`date '+%F %H:%M:%S'`: vrrp transition, `hostname` changed to be $1"
    echo $mailbody | mail -s "$mailsubject" $contact
}
case "$1" in
    master)
        notify master $2
        /etc/rc.d/init.d/haproxy restart
        exit 0
    ;;
    backup)
        notify backup $2 # 在节点切换成backup状态时,无需刻意停止haproxy服务,防止chk_maintaince和chk_haproxy多次对haproxy服务操作;
        exit 0
    ;;
    fault)
        notify fault $2 # 同上
        exit 0
    ;;
    *)
        echo 'Usage: `basename $0` {master|backup|fault}'
        exit 1
    ;;
esac

启动服务

1
service keepalived start # 在两个节点上都需要启动

keepalived双主模型启动

wKiom1Ni8zqzXsjWAAXwJZi3ekQ403.jpg

Haproxy部署

安装配置

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
45
46
47
48
49
50
51
52
53
54
55
yum -y install haproxy # 两节点都需部署
vi /etc/haproxy/haproxy.cfg # 两节点配置一致
global
    log         127.0.0.1 local2
    chroot      /var/lib/haproxy
    pidfile     /var/run/haproxy.pid
    maxconn     4000
    user         haproxy
    group       haproxy
    daemon # 以后台程序运行;
defaults
    mode                   http # 选择HTTP模式,即可进行7层过滤;
    log                     global
    option                  httplog # 可以得到更加丰富的日志输出;
    option                  dontlognull
    option http-server-close # server端可关闭HTTP连接的功能;
    option forwardfor except 127.0.0.0/8 # 传递client端的IP地址给server端,并写入“X-Forward_for”首部中;
    option                  redispatch
    retries                 3
    timeout http-request    10s
    timeout queue           1m
    timeout connect         10s
    timeout client          1m
    timeout server          1m
    timeout http-keep-alive 10s
    timeout check           10s
    maxconn                 30000
listen stats
    mode http
    bind 0.0.0.0:1080 # 统计页面绑定1080端口;
    stats enable # 开启统计页面功能;
    stats hide-version # 隐藏Haproxy版本号;
    stats uri     /haproxyadmin?stats # 自定义统计页面的访问uri;
    stats realm   Haproxy\ Statistics # 统计页面密码验证时的提示信息;
    stats auth    admin:admin # 为统计页面开启登录验证功能;
    stats admin if TRUE # 若登录用户验证通过,则赋予管理功能;
frontend http-in
    bind *:80
    mode http
    log global
    option httpclose
    option logasap
    option dontlognull
    capture request  header Host len 20
    capture request  header Referer len 60
    acl url_static       path_beg       -i /static /images /javascript /stylesheets
    acl url_static       path_end       -i .jpg .jpeg .gif .png .css .js .html
    use_backend static_servers if url_static # 符合ACL规则的,请求转入后端静态服务器
    default_backend dynamic_servers # 默认请求转入后端动态服务器
backend static_servers
    balance roundrobin
    server imgsrv1 192.168.0.25:80 check maxconn 6000 # 静态服务器,可配置多台,还可设置权重weight;
backend dynamic_servers
    balance source # 对于动态请求利用source调度算法,可一定程度上实现session保持;但最好利用cookie绑定的方式实现session保持
    server websrv1 192.168.0.35:80 check maxconn 1000 # 动态服务器,可配置多台,还可设置权重weight;

启动服务

1
service haproxy start # 两节点都需要启动

Nginx部署

见博客“如何测试Nginx的高性能


访问验证

Haproxy统计页面测试

wKioL1Ni9hjzYDblAAV2-8X-c8A579.jpg

wKiom1Ni9A3StmxDABW-x0u1MF0209.jpg

动静分离测试


wKiom1Ni9mOD6XOsAAkk3jT07PM766.jpg

wKioL1Ni9lST0rDAAAj4DLjLm2I967.jpg

高可用测试

wKiom1NjTHXyEohXABCnwqNV-zs908.jpg










本文转自 xxrenzhe11 51CTO博客,原文链接:http://blog.51cto.com/xxrenzhe/1405455,如需转载请自行联系原作者

网友评论

登录后评论
0/500
评论
余二五
+ 关注