Keepalived Httpd的簡單高可用搭建及設定sorry_server

環境：

K1：192.168.11.26

K2：192.168.11.28

R1：192.168.11.21

R2：192.168.11.30

VIP:192.168.11.99

一、K1配置

[root@K1 ~]# cd /etc/keepalived/                    
[root@K1 keepalived]# cat keepalived.conf    #查看配置文件
! Configuration File for keepalived

global_defs {                                #定義全局信息
   notification_email {
	root@localhost
   }
   notification_email_from [email protected]        #定義發送者郵箱，這裏隨意填
   smtp_server 127.0.0.1
   smtp_connect_timeout 30                    #連接超時30秒
   router_id LVS_DEVEL                        #定義id，我們使用默認就好
  # vrrp_mcast_group4 224.0.100.18            #定義組播地址，建議最好使用
}

vrrp_instance VI_1 {                        #定義虛擬路由地址
    state MASTER                            #我們K1爲MASTER
    interface eth0                          #在那口網卡，Centos 7 是enoxxxx
    virtual_router_id 51                    #虛擬路由ID，必須一樣主從
    priority 100                            #優先級
    advert_int 1                            #默認1秒就好，
    authentication {                        #認真
        auth_type PASS                      #keepalived只支持倆中認證，這裏使用第二種簡單密碼認證，也就是預認證
        auth_pass 6Nb6zjWB                  #使用openssh rant -base64 6 獲取隨機密碼，最長支持八位所以也就是後面使用6的原因
    }
    virtual_ipaddress {                     #定義虛擬IP地址
	192.168.11.17
    }
	notify_master "/etc/keepalived/notify.sh master"    #我們直接寫的腳本通知
	notify_backup "/etc/keepalived/notify.sh backup"
	notify_fault "/etc/keepalived/notify.sh fault"
}

virtual_server 192.168.11.17 80 {            #定義keepalived集羣VIP地址
    delay_loop 6                             #檢查週期6秒
    lb_algo rr                               #調度算法
    lb_kind DR                               #調度模型
    protocol TCP                             #只支持TCP，所以想想就好
    sorry_server 127.0.0.1 80                #用於集羣掛掉了，提示的友好界面，比如提示：服務器正在維護等等友好信息...

    real_server 192.168.11.30 80 {           #定義R1
        weight 1                             #權重，我們是rr調度，權重可以忽略
        HTTP_GET {                           #這裏我們使用的HTTP 80端口
            url {                            #定義要請求的url地址
              path /                         #請求地址爲/ ，也就是/indexx.html或者index.php等等
              status_code 200                #返回狀態碼爲200表示OK
            }
            connect_timeout 3                #連接超時時長，3秒
            nb_get_retry 3                   #重試3次
            delay_before_retry 3             #每次重試的時間間隔3秒
        }                                
    }
   real_server 192.168.11.21 80 {            #定義R2
        weight 1                             #權重
        HTTP_GET {                           #這裏我們使用的HTTP 80端口
            url {                            #定義要請求的url地址
              path /                         #請求地址爲/ ，也就是/indexx.html或者index.php等等
              status_code 200                #返回狀態碼爲200表示OK
            }
            connect_timeout 3                #連接超時時長，3秒
            nb_get_retry 3                   #重試3次
            delay_before_retry 3             #每次重試的時間間隔3秒
        }
    }

}

二、K2配置

[root@K2 keepalived]# cat keepalived.conf 
! Configuration File for keepalived

global_defs {
   notification_email {
	root@localhost
   }
   notification_email_from [email protected]
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id LVS_DEVEL
  # vrrp_mcast_group4 224.0.100.18
}

vrrp_instance VI_1 {
    state BACKUP                                            #這裏不同，是BACKUP
    interface eth0
    virtual_router_id 51
    priority 98                                             #這裏不同，優先級是98，下面其他都一樣
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 6Nb6zjWB
    }
    virtual_ipaddress {
	192.168.11.17
    }
	notify_master "/etc/keepalived/notify.sh master"
	notify_backup "/etc/keepalived/notify.sh backup"
	notify_fault "/etc/keepalived/notify.sh fault"
}

virtual_server 192.168.11.17 80 {
    delay_loop 6
    lb_algo rr
    lb_kind DR
    protocol TCP
    sorry_server 127.0.0.1 80

    real_server 192.168.11.30 80 {
        weight 1
        HTTP_GET {
            url {
              path /
              status_code 200
            }
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        }
    }
   real_server 192.168.11.21 80 {
        weight 1
        HTTP_GET {
            url {
              path /
              status_code 200
            }
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        }
    }

}

三、R2配置

[root@R2 ~]# cat skp.sh                                                     #查看腳本內容
#!/bin/bash
#
vip=192.168.11.17
case $1 in
start)
echo 1 > /proc/sys/net/ipv4/conf/all/arp_ignore
echo 1 > /proc/sys/net/ipv4/conf/lo/arp_ignore
echo 2 > /proc/sys/net/ipv4/conf/all/arp_announce
echo 2 > /proc/sys/net/ipv4/conf/lo/arp_announce
ifconfig lo:0 $vip netmask 255.255.255.255 broadcast $vip
route add -host $vip dev lo:0
;;
stop)
echo 0 > /proc/sys/net/ipv4/conf/all/arp_ignore
echo 0 > /proc/sys/net/ipv4/conf/lo/arp_ignore
echo 0 > /proc/sys/net/ipv4/conf/all/arp_announce
echo 0 > /proc/sys/net/ipv4/conf/lo/arp_announce
ifconfig lo:0 del $vip
;;
esac
[root@R2 ~]# sh skp.sh start                                                #執行腳本，記得傳參進去start
[root@R2 ~]# ip add sh                                                      #查看，配置成功：inet 192.168.11.17/32 brd 192.168.11.17 scope global lo:0
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
    inet 192.168.11.17/32 brd 192.168.11.17 scope global lo:0
       valid_lft forever preferred_lft forever
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eno16780032: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP qlen 1000
    link/ether 00:0c:29:c8:1b:d5 brd ff:ff:ff:ff:ff:ff
    inet 192.168.11.30/24 brd 192.168.11.255 scope global eno16780032
       valid_lft forever preferred_lft forever
    inet6 fe80::20c:29ff:fec8:1bd5/64 scope link 
       valid_lft forever preferred_lft forever
[root@R2 ~]# yum install httpd -y                                           #安裝httpd服務
[root@R2 ~]# cat /var/www/html/index.html                                   #標記index.html內容，以做區別，在生產機千萬要別這樣區別，不然用戶會迷茫的！~
<h1>===30===</h1>
[root@R2 ~]# systemctl start httpd                                          #啓動httpd服務，這臺是Centos 7 啓動方式可能有所不同

四、R1配置

[root@R1 ~]# yum install httpd -y                                           #安裝httpd服務
[root@R1 ~]# sh skp.sh start                                                #執行腳本，記得傳參進去
[root@R1 ~]# ip add sh                                                      #查看，配置成功：inet 192.168.11.17/32 brd 192.168.11.17 scope global lo:0
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
    inet 192.168.11.17/32 brd 192.168.11.17 scope global lo:0
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:4e:ce:b9 brd ff:ff:ff:ff:ff:ff
    inet 192.168.11.21/24 brd 192.168.11.255 scope global eth0
    inet6 fe80::20c:29ff:fe4e:ceb9/64 scope link 
       valid_lft forever preferred_lft forever
[root@R1 ~]# cat /var/www/html/index.html                                   #標記index.html內容，以做區別，在生產機千萬要別這樣區別，不然用戶會迷茫的！~
<h1>===21===</h1>
[root@R1 ~]# service httpd start                                            #啓動服務，這是Centos 6 很熟悉吧 ^-^

五、在K2啓動，這臺是初始是BACKUP優先級是98

[root@K2 keepalived]# service keepalived start                #啓動keepalived，提示OK啓動成功
Starting keepalived:                                       [  OK  ]
[root@K2 keepalived]# ipvsadm -Ln                      #查看是否成功，發現是OK的
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.11.17:80 rr
  -> 192.168.11.21:80             Route   1      0          0         
  -> 192.168.11.30:80             Route   1      0          0 
[root@K2 ~]# tail -f /var/log/messages                   #這是在啓動前，在另外的一個session執行的命令，有興趣可以看看哦，後面就示例啦                   
Oct 21 22:58:34 redis_master Keepalived[22646]: Starting Keepalived v1.2.13 (03/19,2015)
Oct 21 22:58:34 redis_master Keepalived[22647]: Starting Healthcheck child process, pid=22648
Oct 21 22:58:34 redis_master Keepalived_healthcheckers[22648]: Netlink reflector reports IP 192.168.11.28 added
Oct 21 22:58:34 redis_master Keepalived[22647]: Starting VRRP child process, pid=22649
Oct 21 22:58:34 redis_master Keepalived_healthcheckers[22648]: Netlink reflector reports IP fe80::20c:29ff:fe2d:bab0 added
Oct 21 22:58:34 redis_master Keepalived_healthcheckers[22648]: Registering Kernel netlink reflector
Oct 21 22:58:34 redis_master Keepalived_healthcheckers[22648]: Registering Kernel netlink command channel
Oct 21 22:58:34 redis_master Keepalived_vrrp[22649]: Netlink reflector reports IP 192.168.11.28 added
Oct 21 22:58:34 redis_master Keepalived_vrrp[22649]: Netlink reflector reports IP fe80::20c:29ff:fe2d:bab0 added
Oct 21 22:58:34 redis_master Keepalived_vrrp[22649]: Registering Kernel netlink reflector
Oct 21 22:58:34 redis_master Keepalived_vrrp[22649]: Registering Kernel netlink command channel
Oct 21 22:58:34 redis_master Keepalived_vrrp[22649]: Registering gratuitous ARP shared channel
Oct 21 22:58:34 redis_master Keepalived_vrrp[22649]: Opening file '/etc/keepalived/keepalived.conf'.
Oct 21 22:58:34 redis_master Keepalived_healthcheckers[22648]: Opening file '/etc/keepalived/keepalived.conf'.
Oct 21 22:58:34 redis_master Keepalived_vrrp[22649]: Configuration is using : 63319 Bytes
Oct 21 22:58:34 redis_master Keepalived_vrrp[22649]: Using LinkWatch kernel netlink reflector...
Oct 21 22:58:34 redis_master Keepalived_healthcheckers[22648]: Configuration is using : 17292 Bytes
Oct 21 22:58:34 redis_master Keepalived_vrrp[22649]: VRRP_Instance(VI_1) Entering BACKUP STATE
Oct 21 22:58:34 redis_master Keepalived_vrrp[22649]: VRRP sockpool: [ifindex(2), proto(112), unicast(0), fd(10,11)]
Oct 21 22:58:34 redis_master Keepalived_healthcheckers[22648]: Using LinkWatch kernel netlink reflector...
Oct 21 22:58:34 redis_master Keepalived_healthcheckers[22648]: Activating healthchecker for service [192.168.11.30]:80
Oct 21 22:58:34 redis_master Keepalived_healthcheckers[22648]: Activating healthchecker for service [192.168.11.21]:80
Oct 21 22:58:38 redis_master Keepalived_vrrp[22649]: VRRP_Instance(VI_1) Transition to MASTER STATE
Oct 21 22:58:39 redis_master Keepalived_vrrp[22649]: VRRP_Instance(VI_1) Entering MASTER STATE
Oct 21 22:58:39 redis_master Keepalived_vrrp[22649]: VRRP_Instance(VI_1) setting protocol VIPs.
Oct 21 22:58:39 redis_master Keepalived_vrrp[22649]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 192.168.11.17
Oct 21 22:58:39 redis_master Keepalived_healthcheckers[22648]: Netlink reflector reports IP 192.168.11.17 added
Oct 21 22:58:44 redis_master Keepalived_vrrp[22649]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 192.168.11.17

[root@localhost ~]# curl http://192.168.11.17                #使用其他的機制訪問，OK也是我們所設定的rr調度算法
<h1>===21===</h1>
[root@localhost ~]# curl http://192.168.11.17
<h1>===30===</h1>
[root@localhost ~]# curl http://192.168.11.17
<h1>===21===</h1>
[root@localhost ~]# curl http://192.168.11.17
<h1>===30===</h1>

六、啓動K1，由於是MASTER 優先級是100 所以VIP地址資源會被搶佔到本機

[root@K1 keepalived]# ip add sh                              #我們發現，VIP地址被搶過來了
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
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:ad:98:5c brd ff:ff:ff:ff:ff:ff
    inet 192.168.11.26/24 brd 192.168.11.255 scope global eth0
    inet 192.168.11.17/32 scope global eth0
    inet6 fe80::20c:29ff:fead:985c/64 scope link 
       valid_lft forever preferred_lft forever
[root@K1 keepalived]# ipvsadm -Ln                            #服務也是正常的，記得K2使用ipvsadm -Ln 看到的也是一樣的
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.11.17:80 rr
  -> 192.168.11.21:80             Route   1      0          0         
  -> 192.168.11.30:80             Route   1      0          0  
[root@localhost ~]# curl http://192.168.11.17                #使用其他的機制訪問，OK也是我們所設定的rr調度算法
<h1>===21===</h1>
[root@localhost ~]# curl http://192.168.11.17
<h1>===30===</h1>
[root@localhost ~]# curl http://192.168.11.17
<h1>===21===</h1>
[root@localhost ~]# curl http://192.168.11.17
<h1>===30===</h1>

七、停掉R1/R2服務，看看是否會是我們所希望的那樣提示sorry_server的友好界面

[root@K1 keepalived]# cat /var/www/html/index.html      #K1的友好界面提示，爲了實驗區分，特地+K1
<h1>===localhost--K1==</h1>
[root@K2 keepalived]# cat /var/www/html/index.html      #K2的友好界面提示，爲了實驗區分，特地+K2
<h1>===localhost--K2===</h1>
[root@R2 ~]# systemctl stop httpd               #停掉R2的httpd服務
[root@K1 keepalived]# ipvsadm -Ln               #K1上查看，192.168.11.30:80的被自動去掉了
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.11.17:80 rr
  -> 192.168.11.21:80             Route   1      0          0 
[root@K2 keepalived]# ipvsadm -Ln              #K2上查看，192.168.11.30:80的被自動去掉了，更K1是一樣的
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.11.17:80 rr
  -> 192.168.11.21:80             Route   1      0          0 
[root@R1 ~]# service httpd stop               #停掉R1的httpd服務
Stopping httpd:                                            [  OK  ]
[root@K1 keepalived]# ipvsadm -Ln              #使用了我們的定義的sorry_server
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.11.17:80 rr
  -> 127.0.0.1:80                 Local   1      0          0   
[root@localhost ~]# curl http://192.168.11.17        #如我們所要的結果
<h1>===localhost--K1==</h1>
[root@K1 keepalived]# service keepalived stop        #停掉K1的keepalived
停止 keepalived：                                          [確定]
[root@localhost ~]# curl http://192.168.11.17       #如我們所要的結果，VIP跑去了K2，至此keepalived搭建的簡單httpd高可用就搭建OK啦 ^-^
<h1>===localhost--K2===</h1>

Keepalived Httpd的簡單高可用搭建及設定sorry_server

MySQL view視圖簡單介紹

Nginx 源碼編譯安裝

Centos 7 Git+httpd 基於密碼認證push

Ansible 一鍵配置安裝Keepalived+Nginx作爲前端，httpd+php作爲後端

nginx 優化(突破十萬併發)

Mac下配置sublime實現LaTeX

https://yachay.unat.edu.pe/blog/index.php?comment_area=format_blog&comment_component=blog&comment_co

linux以太網驅動總結