高性能集群软件-----Keepalived 安装配置

(一)keepalived介绍:

Keepalived是Linux下轻量级的高可用的解决方案。Keepalived主要是通过虚拟路由冗余来实现高可用功能,具有部署和使用非常简单,只需一个配置文件即可。它是根据TCP/IP参考模型的第三,第四和第五层交换机检测到每个服务接点的状态,如果某个服务节点出现异常,或者出现故障,keepalived将检测到,并将出现故障的服务节点从集群中剔除,而在故障节点恢复后,keepalived又可以自动将此服务节点重新加入服务器集群中,这些工作全部自动完成,不需要人工干涉,需要人工完成的只是修复出现故障的服务节点。(2台服务器运行Keepalived,一台为主服务器(MASTER),一台为备份服(BACKUP),但是对外表现为一个虚拟IP,主服务器会发送特定的消息给备份服务器,当备份服务器收不到这个消息的时候,即主服务器宕机的时候, 备份服务器就会接管虚拟IP,继续提供服务,从而保证了高可用性)。keepalived采用的是vrrp来实现的。


Keepalived是一个高可用解决方案,主要是用来防止服务器单点发生故障,可以通过和Nginx配合来实现Web服务的高可用。(其实,Keepalived不仅仅可以和Nginx配合,还可以和很多其他服务配合)

Keepalived+Nginx实现高可用的思路:

第一:请求不要直接打到Nginx上,应该先通过Keepalived(这就是所谓虚拟IP,VIP)

第二:Keepalived应该能监控Nginx的生命状态(提供一个用户自定义的脚本,定期检查Nginx进程状态,进行权重变化,,从而实现Nginx故障切换)


(二)安装和配置:

1,编译安装:

1
2
3
4
5
6
7
8
9
[root@localhost install]#cd /tmp/install
[root@localhost install]# yum install gcc gcc-c++ openssl* kernel-devel net-snmp* libnl* -y 
 [root@localhost install]#yum install -y libnfnetlink-devel
[root@localhost install]#wget http://www.keepalived.org/software/keepalived-1.2.19.tar.gz 
###或者直接在官网上下载
[root@localhost install]#tar zxvf keepalived-1.2.19.tar.gz
[root@localhost install]#cd  keepalived-1.2.19
[root@localhost install]#./configure   --prefix=/usr/local/keepalived
[root@localhost install]#make && make install

2,配置:

  (1)环境说明

hostname
PI 安装软件 OS
Master
192.168.122.120 nginx(已安装).keepalived Centos6.5
backup 192.168.122.121 nginx(已安装).keepalived Centos6.5
vip(虚拟IP)
192.168.122.130

备注:依次在192.168.122.120和192.168.122.121进行编译安装如上边的步骤,安装完成之后,会在/usr/local下生成bin、etc、sbin三个文件夹,主要的配置文件在etc目录下,执行文件在sbin目录下


 (2)依次在120和121上操作:

    (a)####keepalive启动时默认会去读/etc/keepalived/keepalived.conf配置文件,但编译安转后,没有自动创建这个文件,而是放到了/usr/local/etc/keepalived/keepalived.conf

为了以后配置方便,我们手动将/usr/local/etc/keepalived/keepalived.conf放到/etc/keepalived/下


1
2
3
[root@localhost install]#mkdir /etc/keepalived/
[root@localhost install]#cp /usr/local/keepalived/etc/keepalived/keepalived.conf /etc/keepalived/
[root@localhost install]#cp /usr/local/keepalived/etc/sysconfig/keepalived  /etc/sysconfig/keepalived

    (b)#####建立服务启动脚本        

1
2
3
4
5
6
[root@localhost install]#cp /usr/local/keepalived/etc/rc.d/init.d/keepalived /etc/init.d/
[root@localhost install]#chmod +x /etc/init.d/keepalived
####下边是centos7.x版本做成服务
[root@localhost install]#cp /tmp/install/keepalived-1.4.0/keepalived/keepalived.service /lib/systemd/system/keepalived.service
[root@localhost install]#systemctl enable keepalived.service
Created symlink from /etc/systemd/system/multi-user.target.wants/keepalived.service to /usr/lib/systemd/system/keepalived.service.

   (c)####更改启动脚本中keepalive启动参数文件的位置,脚本文件路径是/etc/sysconfig/keepalived(如果是通过yum安装,则这个位置就是正确的),而实际文件是在/usr/local/keepalived/etc/sysconfig/keepalived

#sed -i 's#/etc/sysconfig/keepalived#/usr/local/etc/sysconfig/keepalived#g'  /etc/init.d/keepalived

[root@localhost install]#vi /etc/init.d/keepalived

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
#!/bin/sh
#
# Startup script for the Keepalived daemon
#
# processname: keepalived
# pidfile: /var/run/keepalived.pid
# config: /etc/keepalived/keepalived.conf
# chkconfig: - 21 79
# description: Start and stop Keepalived
# Source function library
/etc/rc.d/init.d/functions
# Source configuration file (we set KEEPALIVED_OPTIONS there)
/usr/local/keepalived/etc/sysconfig/keepalived
RETVAL=0
prog="keepalived"
start() {
    echo -n $"Starting $prog: "
    daemon keepalived ${KEEPALIVED_OPTIONS}
    RETVAL=$?
    echo
    [ $RETVAL -eq 0 ] && touch /var/lock/subsys/$prog
}
"/etc/init.d/keepalived" 74L, 1329C

  (d)给/usr/local/keepalived/sbin/keepalived做个软连接,否则会报错

1
2
3
[root@localhost sbin]# /etc/init.d/keepalived start
Starting keepalived: /bin/bash: keepalived: command not found
                                                           [FAILED]
1
2
3
4
[root@localhost sbin]# ln -s /usr/local/keepalived/sbin/keepalived /usr/bin/
[root@localhost sbin]# /etc/init.d/keepalived start
Starting keepalived:                                       [  OK  ]
[root@localhost sbin]#

 

 (3)配置keepalived.conf 主要配置三个地方:global_defs、vrrp_instance、virtual_ip address

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
root@localhost sbin]# vi /etc/keepalived/keepalived.conf 
! Configuration File for keepalived
global_defs {                         #全局设置
   router_id LVS_DEVEL                   #定义路由标识信息,相同局域网唯一,可自定义
}
vrrp_script check_http {                 #首先在vrrp_script区域定义脚本名字和脚本执行的间隔和脚本执行的优先级变更
   script "/opt/nginx_pid.sh"               #然后在实例(vrrp_instance)里面引用,有点类似脚本里面的函数引用一样:先定义,后引用函数名
   interval 2                      #脚本执行间隔时间
   weight 10                                            #利用权重值和优先级进行运算,从而降低主服务优先级使之变为备服务器(建议先忽略)
}
vrrp_instance VI_1 {                       #定义一个vrrp实例名,可自定义
    state MASTER                         #设定初始状态,可以是MASTER或BACKUP,不过当其他节点keepalived启动时会将priority比较大的节点选举为MASTER,因此该项其实没有实质用途。
    interface eth0                       #具有固定ip的网卡,用来接收和发送vrrp包。如果没有设定mcast_src_ip发送多播数据包的地址,那么将使用这个网卡的ip来发送。
    virtual_router_id 51                #虚拟路由的id号,范围为0-255,取值在0-255之间,用来区分多个instance的VRRP组播。相同的VRID为一个组,他将决定多播的MAC地址
    priority 99                        #当前keepalived的优先级,数字越大,优先级越高
    advert_int 1                        #组播信息发送间隔,两个节点设置必须一样,默认为1秒
    authentication {                    #设置节点间验证信息,所以所有节点必须一致
auth_type PASS                                            #验证类型
auth_pass 1111                                             #密码
    }
    track_script {                      #调用上面定义的检测脚本
chk_http_port 
weight 20
     }
    virtual_ipaddress {                #设置vip,如果不指定网卡,那么默认绑定到interface指定的网卡,可以通过添加dev ethx的方式,绑定到指定的网卡。
192.168.122.130
    }
}


同理,在121上进行操作:

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
vi /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
   router_id LVS_DEVEL
}
vrrp_script chk_http_port  {
   script "/root/chk_httpd.sh"
   interval 2
   weight 10
}
vrrp_instance VI_1 {
    state BACKUP#此处为BACKUP
    interface eth0
    virtual_router_id 51
    priority 96#此处比MASTER要小
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    track_script {
        /opt/nginx_pid.sh 
        weight 20
     }
    virtual_ipaddress {
        192.168.122.130
    }
}


(4)在两台服务器上创建定义 vrrp_script check_http的脚本(nginx_pid.sh脚本的作用是当nginx关闭时自动启动nginx)

1
2
3
4
5
6
7
8
9
10
[root@zabbix ~]# cat /opt/nginx_pid.sh 
#!/bin/bash
A=`ps -C nginx --no-header |wc -l`       
if [ $A -eq 0 ];then                    
       /usr/local/nginx/sbin/nginx
       sleep 3
       if [ `ps -C nginx --no-header |wc -l` -eq 0 ];then
           killall keepalived
       fi
fi




 (5)120和121同时启动keepalived

1
2
3
4
5
6
7
8
9
10
11
12
13
14
[root@hadoop1 ~]# /etc/init.d/keepalived start
Starting keepalived:                                       [  OK  ]
[root@hadoop1 ~]# 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 UP qlen 1000
    link/ether 52:54:00:05:7b:31 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.120/24 brd 192.168.122.255 scope global eth0
    inet 192.168.122.130/32 scope global eth0
    inet6 fe80::5054:ff:fe05:7b31/64 scope link 
       valid_lft forever preferred_lft forever
1
2
3
4
5
6
7
8
9
10
11
12
13
[root@hadoop2 ~]# /etc/init.d/keepalived start
Starting keepalived:                                       [  OK  ]
[root@hadoop2 ~]# 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 UP qlen 1000
    link/ether 52:54:00:c6:65:80 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.121/24 brd 192.168.122.255 scope global eth0
    inet6 fe80::5054:ff:fec6:6580/64 scope link 
       valid_lft forever preferred_lft forever

 打开浏览器进行查看:


 (5)验证实验结果:

     实验一:关闭120的keepalived服务看是否切换:

1
2
3
4
5
6
7
8
9
10
11
[root@hadoop1 ~]# 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 UP qlen 1000
    link/ether 52:54:00:05:7b:31 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.120/24 brd 192.168.122.255 scope global eth0
    inet6 fe80::5054:ff:fe05:7b31/64 scope link 
       valid_lft forever preferred_lft forever

在查看下121的IP地址:

1
2
3
4
5
6
7
8
9
10
11
12
13
[root@hadoop2 ~]# 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 UP qlen 1000
    link/ether 52:54:00:c6:65:80 brd ff:ff:ff:ff:ff:ff
    inet 192.168.122.121/24 brd 192.168.122.255 scope global eth0
    inet 192.168.122.130/32 scope global eth0
    inet6 fe80::5054:ff:fec6:6580/64 scope link 
       valid_lft forever preferred_lft forever
[root@hadoop2 ~]#

已成功切换,在打开浏览器:

高性能集群软件-----Keepalived 安装配置


本文转自 lqbyz 51CTO博客,原文链接:http://blog.51cto.com/liqingbiao/1735397


上一篇:github安卓游戏


下一篇:Photoshop打造印有水墨字画效果的竹简方法