keepalived + haproxy 實現web 雙主模型的高可用負載均衡

1、本文的架構圖：

闡述各服務器用途：

   1、haproxy在本構架中實現的是：負載均衡

   2、keepalived實現對haproxy的高可用

   3、apache static 實現靜態頁面的訪問

   4、aoache dynamic實現動態頁面的訪問，圖中有兩個是實現負載均衡的

配置各功能模塊：

   一、配置haproxy和keepalived

驗證：

        1、當一臺keepalived宕機後，VIP會不會轉移到另外一臺服務器

        2、當一臺haproxy服務出故障，VIP會不會轉移到另外一臺服務器

注意：

        那如果keepalived宕機了，haproxy服務還正常運行，我們要不要讓另外一臺服務器把VIP奪過去呢？

        理論上來講：最好不要，但是我們的keepalived中的腳本監控着haproxy的進程，keepalived宕機之後，就無從得知haproxy的健康狀態，也不能決定自己的優先權priority降不降低了。所以，理論上來講最好不要，但是實際中光靠keepalived是做不到的。

配置：

1、給兩臺服務器分別安裝上keepalived

[root@station139 ~]# yum -y install keepalived

2、配置keepalived

[root@node2 ~]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
   notification_email {
        root@localhost    配置服務狀態變化發送郵件到哪個地址
   }
   notification_email_from kaadmin@localhost
   smtp_server 127.0.0.1    給哪個smtp服務器發郵件
   smtp_connect_timeout 30    聯繫上面smtp服務器30秒聯繫不上，就超時
   router_id LVS_DEVEL
}
vrrp_script chk_haproxy {     本腳本是用來檢測該服務器上haproxy服務的健康狀態的
    script "killall -0 haproxy"
    interval 1
    weight -2
}
vrrp_instance VI_1 {
    state MASTER   這太服務器爲主的keepalived
    interface eth0    通過eth0網卡廣播
    virtual_router_id 200  虛擬路由id要改，如果在一個局域網中有多個keepalived集羣
    priority 100   優先級
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 11112222
    }
    track_script {
        chk_haproxy
    }
    virtual_ipaddress {
        192.168.1.200   本機的虛擬IP
   }
    notify_master "/etc/keepalived/notify.sh master" 各不用狀態下運行的腳本
    notify_backup "/etc/keepalived/notify.sh backup"
    notify_fault "/etc/keepalived/notify.sh fault"
}
vrrp_instance VI_2 {   另外一臺主keepalived的從
    state BACKUP
    interface eth0
    virtual_router_id 57
    priority 99   設置要比另外一臺主keepalived的優先級低
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    track_script {
        chk_mantaince_down
   }
    virtual_ipaddress {
        192.168.1.201
   }
}

3、寫keepalived處在不同狀態下所運行的腳本

#!/bin/bash
# Author: MageEdu <[email protected]>
# description: An example of notify script
#
vip=192.168.1.200
contact='root@localhost'
notify() {
    mailsubject="`hostname` to be $1: $vip 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
        /etc/rc.d/init.d/haproxy start
        exit 0
    ;;
    backup)
        notify backup
        /etc/rc.d/init.d/haproxy stop
        exit 0
    ;;
    fault)
        notify fault
        /etc/rc.d/init.d/haproxy stop
        exit 0
    ;;
    *)
        echo 'Usage: `basename $0` {master|backup|fault}'
        exit 1
    ;;
esac
給腳本以執行權限：
chmod  +x  /etc/keepalived/notify.sh

4、配置haproxy

   因爲要實現動靜分離，那麼我們在配置文件中，就要定義動態資源靜態資源轉移到不同的服務上去

[root@node2 ~]# yum -y install haproxy  安裝haproxy
[root@node2 ~]# vim /etc/haproxy/haproxy.cfg
    #
    log         127.0.0.1 local2
    chroot      /var/lib/haproxy
    pidfile     /var/run/haproxy.pid
    maxconn     4000
    user        haproxy
    group       haproxy
    daemon
    # turn on stats unix socket
    stats socket /var/lib/haproxy/stats
#---------------------------------------------------------------------
# common defaults that all the 'listen' and 'backend' sections will
# use if not designated in their block
#---------------------------------------------------------------------
defaults
    mode                    http      指定haproxy工作模式爲http
    log                     global
    option                  httplog
    option                  dontlognull
    option http-server-close       當客戶端超時時，允許服務端斷開連接
    option forwardfor       except 127.0.0.0/8  在http的響應頭部加入forwardfor
    option                  redispatch    #在使用了基於cookie的會話保持的時候，通常加這麼一項，一旦後端某一server宕機時，能夠將其會話重新派發到其它的upstream servers
    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                 3000
#---------------------------------------------------------------------
# main frontend which proxys to the backends
#---------------------------------------------------------------------
frontend  main *:80  前端代理
    acl url_static       path_beg       -i /static /p_w_picpaths /javascript /stylesheets
    acl url_static       path_end       -i .jpg .gif .png .css .js
    acl url_dynamic      path_end       -i .php
    use_backend static          if url_static
    default_backend             dynamic
#---------------------------------------------------------------------
# static backend for serving up p_w_picpaths, stylesheets and such
#---------------------------------------------------------------------
backend static   後端的靜態請求響應
    balance     roundrobin
    server      static 192.168.1.100:80 inter 3000 rise 2 fall 3 check maxconn 5000
#---------------------------------------------------------------------
# round robin balancing between the various backends
#---------------------------------------------------------------------
backend dynamic   後端的動態請求響應
    balance     roundrobin
    server  dynamic1 192.168.1.101:80 inter 3000 rise 2 fall 3 check maxconn 5000
    server  dynamic2 192.168.1.102:80 inter 3000 rise 2 fall 3 check maxconn 5000
listen statistics
        mode http
        bind *:8080
~       stats enable
        stats auth admin:admin
        stats uri /admin?stats   指定URI的訪問路徑
        stats admin if TRUE
        stats hide-version
        stats refresh 5s
        acl allow src 192.168.0.0/24   定義訪問控制列表
        tcp-request content accept if allow
        tcp-request content reject

5、配置另外一臺haproxy服務器

   因爲兩臺服務器的配置大體相同，我們就直接講以上配置好的複製文件和腳本文件都傳到這臺haproxy服務器上，做下修就可以了

[root@node2 ~]# scp /etc/keepalived/keepalived.conf [email protected]:/etc/keepalived/
[email protected]'s password:
keepalived.conf                                                                               100% 4546     4.4KB/s   00:00 
[root@node2 ~]# scp /etc/keepalived/notify.sh [email protected]:/etc/keepalived/
[email protected]'s password:
notify.sh                                                                                     100%  751     0.7KB/s   00:00 
[root@node2 ~]# scp /etc/haproxy/haproxy.cfg [email protected]:/etc/haproxy/
[email protected]'s password:
haproxy.cfg                                                                                   100% 3529     3.5KB/s   00:00

    傳輸完成，接着來配置 /etc/keepalived/keepalived.conf  因爲兩個節點上的/etc/haproxy/haproxy.cfg相同不用更改

    interface eth0
! Configuration File for keepalived
global_defs {
   notification_email {
        root@localhost
   }
   notification_email_from kaadmin@localhost
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id LVS_DEVEL
}
vrrp_script chk_haproxy {
    script "killall -0 haproxy"
    interval 1
    weight -2
}
vrrp_instance VI_1 {
    state BACKUP      這臺把master改成 backup
    interface eth0
    virtual_router_id 200
    priority 99       優先級調的比上一個低
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 11112222
    }
    track_script {
        chk_haproxy
    }
    virtual_ipaddress {
        192.168.1.200
   }
}
vrrp_instance VI_2 {
    state MASTER      本臺的這個要調成MASTER，上個是backup
    interface eth0
    virtual_router_id 57
    priority 100    這個優先級也要高於上個
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        192.168.1.201
   }
    notify_master "/etc/keepalived/notify.sh master"
    notify_backup "/etc/keepalived/notify.sh backup"
    notify_fault "/etc/keepalived/notify.sh fault"
}

---

注意：

notify_master "/etc/keepalived/notify.sh master"

notify_backup "/etc/keepalived/notify.sh backup"

notify_fault "/etc/keepalived/notify.sh fault"

3個狀態分別要執行的腳本，只能放在 MASTER中，原因是：因爲是互爲主從，每個主的都會有個另外一個主的從，如果

把這 “3個狀態執行腳本” 寫入到從的區域中，那麼另外一個主的從狀態就會執行這個腳本，因爲就會停掉所要高可用的

程序，這就造成了，兩個VIP全部轉移到其中一個服務器上去。

---

我們來驗證下，如果keepalived和haproxy分別宕機，vip會不會轉移：

   在兩個節點上都啓動 keepalived和haproxy服務

[root@node2 ~]# service haproxy start
Starting haproxy:                                          [  OK  ]
[root@node2 ~]# service keepalived start
Starting keepalived:                                       [  OK  ]

以下爲正常情況：

keepalived 1：

keepalived 2：

我們來模擬讓第一個haproxy停止掉，再看下，VIP會不會全到 keepalived 2上面去：

[root@node2 ~]# service haproxy stop
Stopping haproxy:                                          [  OK  ]

查看keepalived 1 和 keepalived 2

   看，都過來了。。。。

---

驗證負載均衡很動靜分離

   我們給3個web服務不同的網頁

   1、給apache static一個靜態頁面，來驗證如果請求的不是以 .php結尾的網頁文件都定向到這太服務器上來

   2、給apache dynamic 1 、2 分別兩個 index.php ，實現對動態網頁的負載均衡

我們給apache static 一個符合-i .jpg .gif .png .css .js的網頁，就給個圖片網頁吧

apache static

scp 1.png [email protected]:/var/www/html

apache dynamic 1

vim /var/www/html/index.php
192.168.1.101
<?php
    phpinfo();
?>

 apache dynamic 2

vim /var/www/html/index.php
192.168.1.102
<?php
    phpinfo();
?>

---

1、我們來請求 1.png 結尾的靜態文件

2、我們來請求 .php結尾的頁面

   如此看來，已經對以.php的動態頁面做了負載均衡了

   我們再通過 192.168.1.201 這個虛擬 ip 訪問試試：

由此看來，也實現了雙主模型了，兩個haproxy同時可以服務了。。。

---

3、我們來看看狀態頁面