Linux 高可用(HA)集群之keepalived详解

http://freeloda.blog.51cto.com/2033581/1280962

大纲

一、前言

二、Keepalived 详解

三、环境准备

四、LVS+Keepalived 实现高可用的前端负载均衡器

一、前言

这篇文章是前几篇文章的总结,我们先简单的总结一下我们前面讲解的内容,前面我们讲解了,LVS(负载均衡器)、Heartbeat、Corosync、Pacemaker、Web高可用集群、MySQL高可用集群、DRDB、iscsi、gfs2、cLVM等,唯一没有讲解的就是LVS可用,也就是前端高可用,我们这一篇博文主要讲解内容。在说这个之前我们得和大家讨论一个问题,也是好多博友问的问题。Heartbeat、Corosync、Keepalived这三个集群组件我们到底选哪个好,首先我想说明的是,Heartbeat、Corosync是属于同一类型,Keepalived与Heartbeat、Corosync,根本不是同一类型的。Keepalived使用的vrrp协议方式,虚拟路由冗余协议 (Virtual Router Redundancy Protocol,简称VRRP);Heartbeat或Corosync是基于主机或网络服务的高可用方式;简单的说就是,Keepalived的目的是模拟路由器的高可用,Heartbeat或Corosync的目的是实现Service的高可用。所以一般Keepalived是实现前端高可用,常用的前端高可用的组合有,就是我们常见的LVS+Keepalived、Nginx+Keepalived、HAproxy+Keepalived。而Heartbeat或Corosync是实现服务的高可用,常见的组合有Heartbeat v3(Corosync)+Pacemaker+NFS+Httpd 实现Web服务器的高可用、Heartbeat v3(Corosync)+Pacemaker+NFS+MySQL 实现MySQL服务器的高可用。总结一下,Keepalived中实现轻量级的高可用,一般用于前端高可用,且不需要共享存储,一般常用于两个节点的高可用。而Heartbeat(或Corosync)一般用于服务的高可用,且需要共享存储,一般用于多节点的高可用。这个问题我们说明白了,又有博友会问了,那heartbaet与corosync我们又应该选择哪个好啊,我想说我们一般用corosync,因为corosync的运行机制更优于heartbeat,就连从heartbeat分离出来的pacemaker都说在以后的开发当中更倾向于corosync,所以现在corosync+pacemaker是最佳组合。但说实话我对于软件没有任何倾向性,所以我把所有的集群软件都和大家说了一下,我认为不管什么软件,只要它能存活下来都有它的特点和应用领域,只有把特定的软件放在特定的位置才能发挥最大的作用,那首先我们得对这个软件有所有了解。学习一种软件的最好方法,就是去查官方文档。好了说了那么多希望大家有所收获,下面我们来说一说keepalived。

二、Keepalived 详解

1.Keepalived 定义

Keepalived 是一个基于VRRP协议来实现的LVS服务高可用方案,可以利用其来避免单点故障。一个LVS服务会有2台服务器运行Keepalived,一台为主服务器(MASTER),一台为备份服务器(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是VRRP的完美实现,因此在介绍keepalived之前,先介绍一下VRRP的原理。

2.VRRP 协议简介

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。

3.VRRP 工作机制

在一个VRRP虚拟路由器中,有多台物理的VRRP路由器,但是这多台的物理的机器并不能同时工作,而是由一台称为MASTER的负责路由工作,其它的都是BACKUP,MASTER并非一成不变,VRRP让每个VRRP路由器参与竞选,最终获胜的就是MASTER。MASTER拥有一些特权,比如,拥有虚拟路由器的IP地址,我们的主机就是用这个IP地址作为静态路由的。拥有特权的MASTER要负责转发发送给网关地址的包和响应ARP请求。

VRRP通过竞选协议来实现虚拟路由器的功能,所有的协议报文都是通过IP多播(multicast)包(多播地址224.0.0.18)形式发送的。虚拟路由器由VRID(范围0-255)和一组IP地址组成,对外表现为一个周知的MAC地址。所以,在一个虚拟路由 器中,不管谁是MASTER,对外都是相同的MAC和IP(称之为VIP)。客户端主机并不需要因为MASTER的改变而修改自己的路由配置,对客户端来说,这种主从的切换是透明的。

在一个虚拟路由器中,只有作为MASTER的VRRP路由器会一直发送VRRP通告信息(VRRPAdvertisement message),BACKUP不会抢占MASTER,除非它的优先级(priority)更高。当MASTER不可用时(BACKUP收不到通告信息), 多台BACKUP中优先级最高的这台会被抢占为MASTER。这种抢占是非常快速的(<1s),以保证服务的连续性。由于安全性考虑,VRRP包使用了加密协议进行加密。

4.VRRP 工作流程

(1).初始化:    
路由器启动时,如果路由器的优先级是255(最高优先级,路由器拥有路由器地址),要发送VRRP通告信息,并发送广播ARP信息通告路由器IP地址对应的MAC地址为路由虚拟MAC,设置通告信息定时器准备定时发送VRRP通告信息,转为MASTER状态;否则进入BACKUP状态,设置定时器检查定时检查是否收到MASTER的通告信息。

(2).Master

  • 设置定时通告定时器;
  • 用VRRP虚拟MAC地址响应路由器IP地址的ARP请求;
  • 转发目的MAC是VRRP虚拟MAC的数据包;
  • 如果是虚拟路由器IP的拥有者,将接受目的地址是虚拟路由器IP的数据包,否则丢弃;
  • 当收到shutdown的事件时删除定时通告定时器,发送优先权级为0的通告包,转初始化状态;
  • 如果定时通告定时器超时时,发送VRRP通告信息;
  • 收到VRRP通告信息时,如果优先权为0,发送VRRP通告信息;否则判断数据的优先级是否高于本机,或相等而且实际IP地址大于本地实际IP,设置定时通告定时器,复位主机超时定时器,转BACKUP状态;否则的话,丢弃该通告包;

(3).Backup

  • 设置主机超时定时器;
  • 不能响应针对虚拟路由器IP的ARP请求信息;
  • 丢弃所有目的MAC地址是虚拟路由器MAC地址的数据包;
  • 不接受目的是虚拟路由器IP的所有数据包;
  • 当收到shutdown的事件时删除主机超时定时器,转初始化状态;
  • 主机超时定时器超时的时候,发送VRRP通告信息,广播ARP地址信息,转MASTER状态;
  • 收到VRRP通告信息时,如果优先权为0,表示进入MASTER选举;否则判断数据的优先级是否高于本机,如果高的话承认MASTER有效,复位主机超时定时器;否则的话,丢弃该通告包;

5.ARP查询处理

当内部主机通过ARP查询虚拟路由器IP地址对应的MAC地址时,MASTER路由器回复的MAC地址为虚拟的VRRP的MAC地址,而不是实际网卡的 MAC地址,这样在路由器切换时让内网机器觉察不到;而在路由器重新启动时,不能主动发送本机网卡的实际MAC地址。如果虚拟路由器开启的ARP代理 (proxy_arp)功能,代理的ARP回应也回应VRRP虚拟MAC地址;好了VRRP的简单讲解就到这里,我们下来讲解一下Keepalived的案例。

三、环境准备

1.操作系统

  • CentOS 6.4 X86_64

2.软件版本

  • ipvsadm.x86_64 0:1.25-10.el6
  • keepalived.x86_64 0:1.2.7-3.el6
  • httpd-2.2.15-29.el6.centos.x86_64

3.实验拓扑

4.时间同步

node1:


1

[[email protected] ~]# ntpdate 202.120.2.101

node2:


1

[[email protected] ~]# ntpdate 202.120.2.101

master:


1

[[email protected] ~]# ntpdate 202.120.2.101

slave:


1

[[email protected] ~]# ntpdate 202.120.2.101

5.主机名互相解析

node1:


1

2

3

4

5

[[email protected] ~]# cat /etc/hosts 

127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4  

::1         localhost localhost.localdomain localhost6 localhost6.localdomain6  

192.168.18.201    node1.test.com    node1  

192.168.18.202    node2.test.com    node2

node2:


1

2

3

4

5

[[email protected] ~]# cat /etc/hosts 

127.0.0.1   localhost localhost.localdomain localhost4 localhost4.localdomain4  

::1         localhost localhost.localdomain localhost6 localhost6.localdomain6  

192.168.18.201    node1.test.com    node1  

192.168.18.202    node2.test.com    node2

6.安装yum源

node1:


1

2

[[email protected] ~]# rpm -ivh http://download.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm

[[email protected] ~]# rpm -ivh http://elrepo.org/elrepo-release-6-5.el6.elrepo.noarch.rpm

node2:


1

2

[[email protected] ~]# rpm -ivh http://download.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm

[[email protected] ~]# rpm -ivh http://elrepo.org/elrepo-release-6-5.el6.elrepo.noarch.rpm

master:


1

2

[[email protected] ~]# rpm -ivh http://download.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm

[[email protected] ~]# rpm -ivh http://elrepo.org/elrepo-release-6-5.el6.elrepo.noarch.rpm

slave:


1

2

[[email protected] ~]# rpm -ivh http://download.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rpm

[[email protected] ~]# rpm -ivh http://elrepo.org/elrepo-release-6-5.el6.elrepo.noarch.rpm

四、LVS+Keepalived 实现高可用的前端负载均衡器

node1:

1.安装httpd


1

[[email protected] ~]# yum install -y httpd

2.配置httpd


1

2

[[email protected] ~]# vim /var/www/html/index.html

<h1>RS1.test.com</h1>

3.启动httpd


1

[[email protected] ~]# service httpd start

4.测试

5.设置开机自启动


1

2

3

[[email protected] ~]# chkconfig httpd on  

[[email protected] ~]# chkconfig httpd --list   

httpd              0:关闭    1:关闭    2:启用    3:启用    4:启用    5:启用    6:关闭

6.配置node1


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

[[email protected] ~]# mkdir src  

[[email protected] ~]# cd src/   

[[email protected] src]# vim realserver.sh

#!/bin/bash  

#   

# Script to start LVS DR real server.   

# description: LVS DR real server   

#   

.  /etc/rc.d/init.d/functions

VIP=192.168.18.200 #修改你的VIP  

host=`/bin/hostname`

case "$1" in

start)   

       # Start LVS-DR real server on this machine.   

        /sbin/ifconfig lo down   

        /sbin/ifconfig lo up   

        echo 1 > /proc/sys/net/ipv4/conf/lo/arp_ignore

        echo 2 > /proc/sys/net/ipv4/conf/lo/arp_announce

        echo 1 > /proc/sys/net/ipv4/conf/all/arp_ignore

        echo 2 > /proc/sys/net/ipv4/conf/all/arp_announce

        /sbin/ifconfig lo:0 $VIP broadcast $VIP netmask 255.255.255.255 up  

        /sbin/route add -host $VIP dev lo:0

;;  

stop)

        # Stop LVS-DR real server loopback device(s).  

        /sbin/ifconfig lo:0 down   

        echo 0 > /proc/sys/net/ipv4/conf/lo/arp_ignore

        echo 0 > /proc/sys/net/ipv4/conf/lo/arp_announce

        echo 0 > /proc/sys/net/ipv4/conf/all/arp_ignore

        echo 0 > /proc/sys/net/ipv4/conf/all/arp_announce

;;  

status)

        # Status of LVS-DR real server.  

        islothere=`/sbin/ifconfig lo:0 | grep $VIP`   

        isrothere=`netstat -rn | grep "lo:0" grep $VIP`   

        if [ ! "$islothere" -o ! "isrothere" ];then

            # Either the route or the lo:0 device   

            # not found.   

            echo "LVS-DR real server Stopped."

        else

            echo "LVS-DR real server Running."

        fi

;;   

*)   

            # Invalid entry.   

            echo "$0: Usage: $0 {start|status|stop}"

            exit 1   

;;   

esac

[[email protected] src]# chmod +x realserver.sh  

[[email protected] src]# ll   

总用量 4   

-rwxr-xr-x 1 root root 1485 8月  22 10:18 realserver.sh

[[email protected] src]# ./realserver.sh start

7.查看配置


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

[[email protected] src]# ifconfig  

eth0      Link encap:Ethernet  HWaddr 00:0C:29:11:92:E4

          inet addr:192.168.18.201  Bcast:192.168.18.255  Mask:255.255.255.0   

          inet6 addr: fe80::20c:29ff:fe11:92e4/64 Scope:Link   

          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1   

          RX packets:115061 errors:0 dropped:0 overruns:0 frame:0   

          TX packets:14979 errors:0 dropped:0 overruns:0 carrier:0   

          collisions:0 txqueuelen:1000   

          RX bytes:43448483 (41.4 MiB)  TX bytes:1224926 (1.1 MiB)

lo        Link encap:Local Loopback

          inet addr:127.0.0.1  Mask:255.0.0.0   

          inet6 addr: ::1/128 Scope:Host   

          UP LOOPBACK RUNNING  MTU:16436  Metric:1   

          RX packets:2 errors:0 dropped:0 overruns:0 frame:0   

          TX packets:2 errors:0 dropped:0 overruns:0 carrier:0   

          collisions:0 txqueuelen:0   

          RX bytes:168 (168.0 b)  TX bytes:168 (168.0 b)

lo:0      Link encap:Local Loopback

          inet addr:192.168.18.200  Mask:255.255.255.255   

          UP LOOPBACK RUNNING  MTU:16436  Metric:1

[[email protected] src]# route -n  

Kernel IP routing table   

Destination     Gateway         Genmask         Flags Metric Ref    Use Iface   

192.168.18.200  0.0.0.0         255.255.255.255 UH    0      0        0 lo   

192.168.18.0    0.0.0.0         255.255.255.0   U     0      0        0 eth0   

169.254.0.0     0.0.0.0         255.255.0.0     U     1002   0        0 eth0   

0.0.0.0         192.168.18.254  0.0.0.0         UG    0      0        0 eth0

[[email protected] src]# cat /proc/sys/net/ipv4/conf/lo/arp_ignore  

1   

[[email protected] src]# cat /proc/sys/net/ipv4/conf/lo/arp_announce   

2   

[[email protected] src]# cat /proc/sys/net/ipv4/conf/all/arp_ignore   

1   

[[email protected] src]# cat /proc/sys/net/ipv4/conf/all/arp_announce

2

好了,node1到这里基本配置完成,下面我们来配置node2。

node2:

1.安装httpd


1

[[email protected] ~]# yum install -y httpd

2.配置httpd


1

2

[[email protected] ~]# vim /var/www/html/index.html

<h1>RS2.test.com</h1>

3.启动httpd


1

[[email protected] ~]# service httpd start

4.测试

5.设置开机自启动


1

2

3

[[email protected] ~]# chkconfig httpd on  

[[email protected] ~]# chkconfig httpd --list   

httpd              0:关闭    1:关闭    2:启用    3:启用    4:启用    5:启用    6:关闭

6.配置node2


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

[[email protected] ~]# mkdir src  

[[email protected] ~]# cd src/   

[[email protected] src]# vim realserver.sh

#!/bin/bash  

#   

# Script to start LVS DR real server.   

# description: LVS DR real server   

#   

.  /etc/rc.d/init.d/functions

VIP=192.168.18.200  

host=`/bin/hostname`

case "$1" in

start)   

       # Start LVS-DR real server on this machine.   

        /sbin/ifconfig lo down   

        /sbin/ifconfig lo up   

        echo 1 > /proc/sys/net/ipv4/conf/lo/arp_ignore

        echo 2 > /proc/sys/net/ipv4/conf/lo/arp_announce

        echo 1 > /proc/sys/net/ipv4/conf/all/arp_ignore

        echo 2 > /proc/sys/net/ipv4/conf/all/arp_announce

        /sbin/ifconfig lo:0 $VIP broadcast $VIP netmask 255.255.255.255 up  

        /sbin/route add -host $VIP dev lo:0

;;  

stop)

        # Stop LVS-DR real server loopback device(s).  

        /sbin/ifconfig lo:0 down   

        echo 0 > /proc/sys/net/ipv4/conf/lo/arp_ignore

        echo 0 > /proc/sys/net/ipv4/conf/lo/arp_announce

        echo 0 > /proc/sys/net/ipv4/conf/all/arp_ignore

        echo 0 > /proc/sys/net/ipv4/conf/all/arp_announce

;;  

status)

        # Status of LVS-DR real server.  

        islothere=`/sbin/ifconfig lo:0 | grep $VIP`   

        isrothere=`netstat -rn | grep "lo:0" grep $VIP`   

        if [ ! "$islothere" -o ! "isrothere" ];then

            # Either the route or the lo:0 device   

            # not found.   

            echo "LVS-DR real server Stopped."

        else

            echo "LVS-DR real server Running."

        fi

;;   

*)   

            # Invalid entry.   

            echo "$0: Usage: $0 {start|status|stop}"

            exit 1   

;;   

esac

[[email protected] src]# chmod +x realserver.sh  

[[email protected] src]# ./realserver.sh start

7.查看配置


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

[[email protected] src]# ifconfig  

eth0      Link encap:Ethernet  HWaddr 00:0C:29:B8:DF:6A

          inet addr:192.168.18.202  Bcast:192.168.18.255  Mask:255.255.255.0   

          inet6 addr: fe80::20c:29ff:feb8:df6a/64 Scope:Link   

          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1   

          RX packets:110545 errors:0 dropped:0 overruns:0 frame:0   

          TX packets:10461 errors:0 dropped:0 overruns:0 carrier:0   

          collisions:0 txqueuelen:1000   

          RX bytes:32853789 (31.3 MiB)  TX bytes:889109 (868.2 KiB)

lo        Link encap:Local Loopback

          inet addr:127.0.0.1  Mask:255.0.0.0   

          inet6 addr: ::1/128 Scope:Host   

          UP LOOPBACK RUNNING  MTU:16436  Metric:1   

          RX packets:6 errors:0 dropped:0 overruns:0 frame:0   

          TX packets:6 errors:0 dropped:0 overruns:0 carrier:0   

          collisions:0 txqueuelen:0   

          RX bytes:616 (616.0 b)  TX bytes:616 (616.0 b)

lo:0      Link encap:Local Loopback

          inet addr:192.168.18.200  Mask:255.255.255.255   

          UP LOOPBACK RUNNING  MTU:16436  Metric:1

[[email protected] src]# route -n  

Kernel IP routing table   

Destination     Gateway         Genmask         Flags Metric Ref    Use Iface   

192.168.18.200  0.0.0.0         255.255.255.255 UH    0      0        0 lo   

192.168.18.0    0.0.0.0         255.255.255.0   U     0      0        0 eth0   

169.254.0.0     0.0.0.0         255.255.0.0     U     1002   0        0 eth0   

0.0.0.0         192.168.18.254  0.0.0.0         UG    0      0        0 eth0

[[email protected] src]# cat /proc/sys/net/ipv4/conf/lo/arp_ignore  

1   

[[email protected] src]# cat /proc/sys/net/ipv4/conf/lo/arp_announce   

2   

[[email protected] src]# cat /proc/sys/net/ipv4/conf/all/arp_ignore   

1   

[[email protected] src]# cat /proc/sys/net/ipv4/conf/all/arp_announce   

2

好了,到这里node2也基本配置完成。下面我们来配置master与slave。

masterg与slave:

1.安装keepalived与ipvsadm


1

2

[[email protected] ~]# yum install -y keepalived ipvsadm

[[email protected] ~]# yum install -y keepalived ipvsadm

2.修改配置文件


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

[[email protected] ~]# cat /etc/keepalived/keepalived.conf  

! Configuration File for keepalived

global_defs {  

   notification_email {   

[email protected] #配置管理员邮箱   

   }   

   notification_email_from root #配置发件人   

   smtp_server 127.0.0.1 #配置邮件服务器   

   smtp_connect_timeout 30   

   router_id LVS_DEVEL   

}

vrrp_instance VI_1 {  

    state MASTER #配置模式   

    interface eth0   

    virtual_router_id 51   

    priority 101 #配置优先级   

    advert_int 1   

    authentication {   

        auth_type PASS   

        auth_pass 1111   

    }   

    virtual_ipaddress {   

        192.168.18.200 #配置虚拟IP地址   

    }   

}

virtual_server 192.168.18.200 80 {  

    delay_loop 6   

    lb_algo rr   

    lb_kind DR   

    nat_mask 255.255.255.0   

    #persistence_timeout 50   

    protocol TCP

    real_server 192.168.18.201 80 { #配置realaserver  

        weight 1   

        HTTP_GET { #监控配置   

            url {   

              path /   

          status_code 200   

            }   

            connect_timeout 2   

            nb_get_retry 3   

            delay_before_retry 1   

        }   

    }   

    real_server 192.168.18.202 80 {   

        weight 1   

        HTTP_GET {   

            url {   

              path /   

              status_code 200   

            }   

            connect_timeout 2   

            nb_get_retry 3   

            delay_before_retry 1   

        }   

    }   

}

3.将配置文件同步到slave


1

[[email protected] ~]# scp /etc/keepalived/keepalived.conf [email protected]:/etc/keepalived/

4.简单修改一下slave配置文件


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

[[email protected] keepalived]# cat keepalived.conf  

! Configuration File for keepalived

global_defs {  

   notification_email {   

     [email protected]   

   }   

   notification_email_from root   

   smtp_server 127.0.0.1   

   smtp_connect_timeout 30   

   router_id LVS_DEVEL   

}

vrrp_instance VI_1 {  

    state BACKUP #修改为BACKUP   

    interface eth0   

    virtual_router_id 51   

    priority 100 #修改优先级   

    advert_int 1   

    authentication {   

        auth_type PASS   

        auth_pass 1111   

    }   

    virtual_ipaddress {   

        192.168.18.200   

    }   

}

virtual_server 192.168.18.200 80 {  

    delay_loop 6   

    lb_algo rr   

    lb_kind DR   

    nat_mask 255.255.255.0   

    #persistence_timeout 50   

    protocol TCP

    real_server 192.168.18.201 80 {  

        weight 1   

        HTTP_GET {   

            url {   

              path /   

          status_code 200   

            }   

            connect_timeout 2   

            nb_get_retry 3   

            delay_before_retry 1   

        }   

    }   

    real_server 192.168.18.202 80 {   

        weight 1   

        HTTP_GET {   

            url {   

              path /   

              status_code 200   

            }   

            connect_timeout 2   

            nb_get_retry 3   

            delay_before_retry 1   

        }   

    }   

}

5.启动master与slave的keepalived服务


1

2

3

4

[[email protected] ~]# service keepalived start  

正在启动 keepalived:                                      [确定]

[[email protected] ~]# service keepalived start  

正在启动 keepalived:                                      [确定]

6.查看一下LVS状态


1

2

3

4

5

6

7

[[email protected] ~]# ipvsadm -L -n  

IP Virtual Server version 1.2.1 (size=4096)   

Prot LocalAddress:Port Scheduler Flags   

  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn   

TCP  192.168.18.200:80 rr   

  -> 192.168.18.201:80            Route   1      0          0       

  -> 192.168.18.202:80            Route   1      0          0

7.测试

8.模拟故障

(1).停止一下node1


1

2

[[email protected] src]# service httpd stop  

停止 httpd:                                               [确定]

(2).查看一下的lvs


1

2

3

4

5

6

[[email protected] ~]# ipvsadm -L -n  

IP Virtual Server version 1.2.1 (size=4096)   

Prot LocalAddress:Port Scheduler Flags   

  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn   

TCP  192.168.18.200:80 rr   

  -> 192.168.18.202:80            Route   1      0          0

(3).测试一下

(4).查看一下邮件

(5).重新启动一下node1


1

2

[[email protected] src]# service httpd start  

正在启动 httpd:                                           [确定]

(6).再查看一下lvs状态


1

2

3

4

5

6

7

[[email protected] ~]# ipvsadm -L -n  

IP Virtual Server version 1.2.1 (size=4096)   

Prot LocalAddress:Port Scheduler Flags   

  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn   

TCP  192.168.18.200:80 rr   

  -> 192.168.18.201:80            Route   1      0          0       

  -> 192.168.18.202:80            Route   1      0          0

(7).再查看一下邮件

(8).关闭master上keepalived


1

2

3

4

5

6

[[email protected] ~]# service keepalived stop  

停止 keepalived:                                          [确定]   

[[email protected] ~]# ipvsadm -L -n   

IP Virtual Server version 1.2.1 (size=4096)   

Prot LocalAddress:Port Scheduler Flags   

  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn

(9).查看一下slave状态


1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

[[email protected] ~]# ip addr show  

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 00:0c:29:f9:e6:26 brd ff:ff:ff:ff:ff:ff   

    inet 192.168.18.207/24 brd 192.168.18.255 scope global eth0   

    inet 192.168.18.200/32 scope global eth0   

    inet6 fe80::20c:29ff:fef9:e626/64 scope link   

       valid_lft forever preferred_lft forever   

[[email protected] ~]# ipvsadm -L -n   

IP Virtual Server version 1.2.1 (size=4096)   

Prot LocalAddress:Port Scheduler Flags   

  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn   

TCP  192.168.18.200:80 rr   

  -> 192.168.18.201:80            Route   1      0          0       

  -> 192.168.18.202:80            Route   1      0          0

(10).再次测试一下

注,大家可以看到,经过上面的演示我们现在LVS的高可用即前端负载均衡的高可用,同时实现对后端realserver监控,也实现后端realserver宕机时会给管理员发送邮件。但还有几个问题我们还没有解决,问题如下:

  • 所有realserver都down机,怎么处理?是不是用户就没法打开,还是提供一下维护页面。
  • 怎么完成维护模式keepalived切换?
  • 如何在keepalived故障时,发送警告邮件给指定的管理员?

9.所有realserver都down机,怎么处理?

问题:在集群中如果所有real server全部宕机了,客户端访问时就会出现错误页面,这样是很不友好的,我们得提供一个维护页面来提醒用户,服务器正在维护,什么时间可以访问等,下面我们就来解决一下这个问题。解决方案有两种,一种是提供一台备用的real server当所有的服务器宕机时,提供维护页面,但这样做有点浪费服务器。另一种就是在负载均衡器上提供维护页面,这样是比较靠谱的,也比较常用。下面我们就来具体操作一下。

(1).master与slave安装上httpd


1

2

[[email protected] ~]# yum install -y httpd

[[email protected] ~]# yum install -y httpd

(2).配置维护页面


1

2

3

4

[[email protected] ~]# vim /var/www/html/index.html 

Website is currently under maintenance, please come back later!

 [[email protected] ~]# vim /var/www/html/index.html

Website is currently under maintenance, please come back later!

(3).启动httpd服务并测试


1

2

3

4

[[email protected] ~]# service httpd start 

正在启动 httpd:                                           [确定]

[[email protected] ~]# service httpd start 

正在启动 httpd:                                           [确定]

(4).修改配置文件

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

45

46

47

48

49

50

51

52

53

54

55

56

57

[[email protected] ~]# cat /etc/keepalived/keepalived.conf  

! Configuration File for keepalived

global_defs { 

   notification_email {  

     [email protected]  

   }  

   notification_email_from root  

   smtp_server 127.0.0.1  

   smtp_connect_timeout 30  

   router_id LVS_DEVEL  

}

vrrp_instance VI_1 { 

    state MASTER  

    interface eth0  

    virtual_router_id 51  

    priority 101  

    advert_int 1  

    authentication {  

        auth_type PASS  

        auth_pass 1111  

    }  

    virtual_ipaddress {  

        192.168.18.200  

    }  

}

virtual_server 192.168.18.200 80 { 

    delay_loop 6  

    lb_algo rr  

    lb_kind DR  

    nat_mask 255.255.255.0  

    protocol TCP

    real_server 192.168.18.201 80 { 

        weight 1  

        HTTP_GET {  

            url {  

              path /  

          status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

    real_server 192.168.18.202 80 {  

        weight 1  

        HTTP_GET {  

            url {  

              path /  

              status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

   sorry_server 127.0.0.1 80   #增加一行sorry_server   

}

slave:


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

[[email protected] ~]# cat /etc/keepalived/keepalived.conf  

! Configuration File for keepalived

global_defs { 

   notification_email {  

     [email protected]  

   }  

   notification_email_from root  

   smtp_server 127.0.0.1  

   smtp_connect_timeout 30  

   router_id LVS_DEVEL  

}

vrrp_instance VI_1 { 

    state BACKUP  

    interface eth0  

    virtual_router_id 51  

    priority 100  

    advert_int 1  

    authentication {  

        auth_type PASS  

        auth_pass 1111  

    }  

    virtual_ipaddress {  

        192.168.18.200  

    }  

}

virtual_server 192.168.18.200 80 { 

    delay_loop 6  

    lb_algo rr  

    lb_kind DR  

    nat_mask 255.255.255.0  

    protocol TCP

    real_server 192.168.18.201 80 { 

        weight 1  

        HTTP_GET {  

            url {  

              path /  

          status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

    real_server 192.168.18.202 80 {  

        weight 1  

        HTTP_GET {  

            url {  

              path /  

              status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

    sorry_server 127.0.0.1 80 #增加一行sorry_server   

}

(5).关闭所有的real server并重新启动一下master与slave的keepalived


1

2

3

4

5

6

7

8

9

10

[[email protected] ~]# service httpd stop 

停止 httpd:                                               [确定]

[[email protected] ~]# service httpd stop 

停止 httpd:                                               [确定]

[[email protected]er ~]# service keepalived restart 

停止 keepalived:                                          [确定]  

正在启动 keepalived:                                      [确定]

[[email protected] ~]# service keepalived restart 

停止 keepalived:                                          [确定]  

正在启动 keepalived:                                      [确定]

(6).查看一下lvs


1

2

3

4

5

6

[[email protected] ~]# ipvsadm -L -n 

IP Virtual Server version 1.2.1 (size=4096)  

Prot LocalAddress:Port Scheduler Flags  

  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn  

TCP  192.168.18.200:80 rr  

  -> 127.0.0.1:80                 Local   1      0          0

(7).测试

注,sorry_server测试成功,下面我们继续。

10.怎么完成维护模式keepalived切换?

问题:我们一般进行主从切换测试时都是关闭keepalived或关闭网卡接口,有没有一种方法能实现在不关闭keepalived下或网卡接口来实现维护呢?方法肯定是有的,在keepalived新版本中,支持脚本vrrp_srcipt,具体如何使用大家可以man keepalived.conf查看。下面我们来演示一下具体怎么实现。

(1).定义脚本


1

2

3

4

5

6

7

vrrp_srcipt chk_schedown { 

   script "[ -e /etc/keepalived/down ] && exit 1 || exit 0"

   interval 1 #监控间隔  

   weight -5 #减小优先级  

   fall 2 #监控失败次数  

   rise 1 #监控成功次数  

}

(2).执行脚本


1

2

3

track_script {

   chk_schedown #执行chk_schedown脚本

}

(3).修改配置文件

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

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

[[email protected] ~]# cat /etc/keepalived/keepalived.conf  

! Configuration File for keepalived

global_defs { 

   notification_email {  

     [email protected]  

   }  

   notification_email_from root  

   smtp_server 127.0.0.1  

   smtp_connect_timeout 30  

   router_id LVS_DEVEL  

}

vrrp_script chk_schedown { #定义vrrp执行脚本  

   script "[ -e /etc/keepalived/down ] && exit 1 || exit 0" #查看是否有down文件,有就进入维护模式  

   interval 1 #监控间隔时间  

   weight -5 #降低优先级  

   fall 2 #失败次数  

   rise 1 #成功数次  

}

vrrp_instance VI_1 { 

    state MASTER  

    interface eth0  

    virtual_router_id 51  

    priority 101  

    advert_int 1  

    authentication {  

        auth_type PASS  

        auth_pass 1111  

    }  

    virtual_ipaddress {  

        192.168.18.200  

    }  

    track_script { #执行脚本  

        chk_schedown  

    }  

}

virtual_server 192.168.18.200 80 { 

    delay_loop 6  

    lb_algo rr  

    lb_kind DR  

    nat_mask 255.255.255.0  

    protocol TCP

    real_server 192.168.18.201 80 { 

        weight 1  

        HTTP_GET {  

            url {  

              path /  

          status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

    real_server 192.168.18.202 80 {  

        weight 1  

        HTTP_GET {  

            url {  

              path /  

              status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

   sorry_server 127.0.0.1 80  

}

slave:


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

[[email protected] ~]# cat /etc/keepalived/keepalived.conf  

! Configuration File for keepalived

global_defs { 

   notification_email {  

     [email protected]  

   }  

   notification_email_from root  

   smtp_server 127.0.0.1  

   smtp_connect_timeout 30  

   router_id LVS_DEVEL  

}

vrrp_script chk_schedown { 

   script "[ -e /etc/keepalived/down ] && exit 1 || exit 0"

   interval 1  

   weight -5  

   fall 2  

   rise 1  

}

vrrp_instance VI_1 { 

    state BACKUP  

    interface eth0  

    virtual_router_id 51  

    priority 100  

    advert_int 1  

    authentication {  

        auth_type PASS  

        auth_pass 1111  

    }  

    virtual_ipaddress {  

        192.168.18.200  

    }  

    track_script {  

    chk_schedown  

    }

}

virtual_server 192.168.18.200 80 { 

    delay_loop 6  

    lb_algo rr  

    lb_kind DR  

    nat_mask 255.255.255.0  

    protocol TCP

    real_server 192.168.18.201 80 { 

        weight 1  

        HTTP_GET {  

            url {  

              path /  

          status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

    real_server 192.168.18.202 80 {  

        weight 1  

        HTTP_GET {  

            url {  

              path /  

              status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

    sorry_server 127.0.0.1 80

(4).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

[[email protected] keepalived]# touch down #新建一下down文件

[[email protected] keepalived]# ll 

总用量 4  

-rw-r--r-- 1 root root    0 8月  22 13:39 down  

-rw-r--r-- 1 root root 1317 8月  22 13:35 keepalived.conf

[[email protected] keepalived]# tail -f /var/log/messages #查看一下日志 

Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) Entering MASTER STATE  

Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) setting protocol VIPs.  

Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) Sending gratuitous ARPs on eth0 for 192.168.18.200  

Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) Received higher prio advert  

Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) Entering BACKUP STATE  

Aug 22 13:43:52 master Keepalived_vrrp[12003]: VRRP_Instance(VI_1) removing protocol VIPs.  

Aug 22 13:43:52 master Keepalived_healthcheckers[12002]: Netlink reflector reports IP 192.168.18.200 added  

Aug 22 13:43:52 master Keepalived_healthcheckers[12002]: Netlink reflector reports IP 192.168.18.200 removed  

Aug 22 13:43:52 master Keepalived_healthcheckers[12002]: SMTP alert successfully sent.  

Aug 22 13:43:52 master Keepalived_healthcheckers[12002]: SMTP alert successfully sent.  

^C  

[[email protected] keepalived]# ip add show #查看VIP  

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 00:0c:29:4b:a1:85 brd ff:ff:ff:ff:ff:ff  

    inet 192.168.18.208/24 brd 192.168.18.255 scope global eth0  

    inet6 fe80::20c:29ff:fe4b:a185/64 scope link   

       valid_lft forever preferred_lft forever

slave:


1

2

3

4

5

6

7

8

9

10

11

12

[[email protected] ~]# ip addr show #查看一下VIP已转移到slave上 

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 00:0c:29:f9:e6:26 brd ff:ff:ff:ff:ff:ff  

    inet 192.168.18.207/24 brd 192.168.18.255 scope global eth0  

    inet 192.168.18.200/32 scope global eth0  

    inet6 fe80::20c:29ff:fef9:e626/64 scope link   

       valid_lft forever preferred_lft forever

好了,自写监测脚本,完成维护模式切换,到这里就演示成功,下面我们来解决最后一个问题,就是keepalived主从切换的邮件通告。

11.如何在keepalived故障时(或主备切换时),发送警告邮件给指定的管理员?

(1).keepalived通知脚本进阶示例

下面的脚本可以接受选项,其中

  • -s, --service SERVICE,...:指定服务脚本名称,当状态切换时可自动启动、重启或关闭此服务;
  • -a, --address VIP: 指定相关虚拟路由器的VIP地址;
  • -m, --mode {mm|mb}:指定虚拟路由的模型,mm表示主主,mb表示主备;它们表示相对于同一种服务而方,其VIP的工作类型;
  • -n, --notify {master|backup|fault}:指定通知的类型,即vrrp角色切换的目标角色;
  • -h, --help:获取脚本的使用帮助;

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

#!/bin/bash 

# Author: freeloda   

# description: An example of notify script  

# Usage: notify.sh -m|--mode {mm|mb} -s|--service SERVICE1,... -a|--address VIP  -n|--notify {master|backup|falut} -h|--help

contact=‘[email protected]‘

helpflag=0  

serviceflag=0  

modeflag=0  

addressflag=0  

notifyflag=0

Usage() { 

  echo "Usage: notify.sh [-m|--mode {mm|mb}] [-s|--service SERVICE1,...] <-a|--address VIP>  <-n|--notify {master|backup|falut}>"

  echo "Usage: notify.sh -h|--help"

}

ParseOptions() { 

  local I=1;  

  if [ $# -gt 0 ]; then  

    while [ $I -le $# ]; do  

      case $1 in

      -s|--service)  

        [ $# -lt 2 ] && return 3  

         serviceflag=1  

         services=(`echo $2|awk -F"," ‘{for(i=1;i<=NF;i++) print $i}‘`)  

        shift 2 ;;  

      -h|--help)  

         helpflag=1  

        return 0  

        shift

        ;;  

      -a|--address)  

        [ $# -lt 2 ] && return 3  

        addressflag=1  

        vip=$2  

        shift 2  

        ;;  

      -m|--mode)  

        [ $# -lt 2 ] && return 3  

        mode=$2  

        shift 2  

        ;;  

      -n|--notify)  

        [ $# -lt 2 ] && return 3  

        notifyflag=1  

        notify=$2  

        shift 2  

        ;;  

      *)  

        echo "Wrong options..."

        Usage  

        return 7  

        ;;  

       esac

    done

    return 0  

  fi

}

#workspace=$(dirname $0)

RestartService() { 

  if [ ${#@} -gt 0 ]; then  

    for in [email protected]; do

      if [ -x /etc/rc.d/init.d/$I ]; then

        /etc/rc.d/init.d/$I restart  

      else

        echo "$I is not a valid service..."

      fi

    done

  fi

}

StopService() { 

  if [ ${#@} -gt 0 ]; then  

    for in [email protected]; do

      if [ -x /etc/rc.d/init.d/$I ]; then

        /etc/rc.d/init.d/$I stop  

      else

        echo "$I is not a valid service..."

      fi

    done

  fi

}

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  

}

# Main Function  

ParseOptions [email protected]  

[ $? -ne 0 ] && Usage && exit 5

[ $helpflag -eq 1 ] && Usage && exit 0

if [ $addressflag -ne 1 -o $notifyflag -ne 1 ]; then

  Usage  

  exit 2  

fi

mode=${mode:-mb}

case $notify in

‘master‘)  

  if [ $serviceflag -eq 1 ]; then

      RestartService ${services[*]}  

  fi

  Notify master  

  ;;  

‘backup‘)  

  if [ $serviceflag -eq 1 ]; then

    if "$mode" == ‘mb‘ ]; then

      StopService ${services[*]}  

    else

      RestartService ${services[*]}  

    fi

  fi

  Notify backup  

  ;;  

‘fault‘)  

  Notify fault  

  ;;  

*)  

  Usage  

  exit 4  

  ;;  

esac

(2).在keepalived.conf配置文件中,其调用方法如下所示:

  • notify_master "/etc/keepalived/notify.sh -n master -a 192.168.18.200"
  • notify_backup "/etc/keepalived/notify.sh -n backup -a 192.168.18.200"
  • notify_fault "/etc/keepalived/notify.sh -n fault -a 192.168.18.200"

(3).修改配置文件

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

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

[[email protected] keepalived]# cat keepalived.conf  

! Configuration File for keepalived

global_defs { 

   notification_email {  

     [email protected]  

   }  

   notification_email_from root  

   smtp_server 127.0.0.1  

   smtp_connect_timeout 30  

   router_id LVS_DEVEL  

}

vrrp_script chk_schedown { 

   script "[ -e /etc/keepalived/down ] && exit 1 || exit 0"

   interval 1  

   weight -5  

   fall 2  

   rise 1  

}

vrrp_instance VI_1 { 

    state MASTER  

    interface eth0  

    virtual_router_id 51  

    priority 101  

    advert_int 1  

    authentication {  

        auth_type PASS  

        auth_pass 1111  

    }  

    virtual_ipaddress {  

        192.168.18.200  

    }  

    track_script {  

        chk_schedown  

    }

    #增加以下三行 

    notify_master "/etc/keepalived/notify.sh -n master -a 192.168.18.200"

    notify_backup "/etc/keepalived/notify.sh -n backup -a 192.168.18.200"

    notify_fault "/etc/keepalived/notify.sh -n fault -a 192.168.18.200"

}

virtual_server 192.168.18.200 80 { 

    delay_loop 6  

    lb_algo rr  

    lb_kind DR  

    nat_mask 255.255.255.0  

    protocol TCP

    real_server 192.168.18.201 80 { 

        weight 1  

        HTTP_GET {  

            url {  

              path /  

          status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

    real_server 192.168.18.202 80 {  

        weight 1  

        HTTP_GET {  

            url {  

              path /  

              status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

   sorry_server 127.0.0.1 80  

}

slave:


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

[[email protected] keepalived]# cat keepalived.conf  

! Configuration File for keepalived

global_defs { 

   notification_email {  

     [email protected]  

   }  

   notification_email_from root  

   smtp_server 127.0.0.1  

   smtp_connect_timeout 30  

   router_id LVS_DEVEL  

}

vrrp_script chk_schedown { 

   script "[ -e /etc/keepalived/down ] && exit 1 || exit 0"

   interval 1  

   weight -5  

   fall 2  

   rise 1  

}

vrrp_instance VI_1 { 

    state BACKUP  

    interface eth0  

    virtual_router_id 51  

    priority 100  

    advert_int 1  

    authentication {  

        auth_type PASS  

        auth_pass 1111  

    }  

    virtual_ipaddress {  

        192.168.18.200  

    }  

    track_script {  

    chk_schedown  

    }

    #增加以下三行 

    notify_master "/etc/keepalived/notify.sh -n master -a 192.168.18.200"

    notify_backup "/etc/keepalived/notify.sh -n backup -a 192.168.18.200"

    notify_fault "/etc/keepalived/notify.sh -n fault -a 192.168.18.200"

}

virtual_server 192.168.18.200 80 { 

    delay_loop 6  

    lb_algo rr  

    lb_kind DR  

    nat_mask 255.255.255.0  

    protocol TCP

    real_server 192.168.18.201 80 { 

        weight 1  

        HTTP_GET {  

            url {  

              path /  

          status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

    real_server 192.168.18.202 80 {  

        weight 1  

        HTTP_GET {  

            url {  

              path /  

              status_code 200  

            }  

            connect_timeout 2  

            nb_get_retry 3  

            delay_before_retry 1  

        }  

    }  

    sorry_server 127.0.0.1 80  

}

(4).增加脚本


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

[[email protected] keepalived]# pwd 

/etc/keepalived

[[email protected] keepalived]# vim notify.sh

[[email protected] keepalived]# cat notify.sh  

#!/bin/bash  

# Author: freeloda   

# description: An example of notify script  

# Usage: notify.sh -m|--mode {mm|mb} -s|--service SERVICE1,... -a|--address VIP  -n|--notify {master|backup|falut} -h|--help

contact=‘[email protected]‘

helpflag=0  

serviceflag=0  

modeflag=0  

addressflag=0  

notifyflag=0

Usage() { 

  echo "Usage: notify.sh [-m|--mode {mm|mb}] [-s|--service SERVICE1,...] <-a|--address VIP>  <-n|--notify {master|backup|falut}>"

  echo "Usage: notify.sh -h|--help"

}

ParseOptions() { 

  local I=1;  

  if [ $# -gt 0 ]; then  

    while [ $I -le $# ]; do  

      case $1 in

      -s|--service)  

        [ $# -lt 2 ] && return 3  

         serviceflag=1  

         services=(`echo $2|awk -F"," ‘{for(i=1;i<=NF;i++) print $i}‘`)  

        shift 2 ;;  

      -h|--help)  

         helpflag=1  

        return 0  

        shift

        ;;  

      -a|--address)  

        [ $# -lt 2 ] && return 3  

        addressflag=1  

        vip=$2  

        shift 2  

        ;;  

      -m|--mode)  

        [ $# -lt 2 ] && return 3  

        mode=$2  

        shift 2  

        ;;  

      -n|--notify)  

        [ $# -lt 2 ] && return 3  

        notifyflag=1  

        notify=$2  

        shift 2  

        ;;  

      *)  

        echo "Wrong options..."

        Usage  

        return 7  

        ;;  

       esac

    done

    return 0  

  fi

}

#workspace=$(dirname $0)

RestartService() { 

  if [ ${#@} -gt 0 ]; then  

    for in [email protected]; do

      if [ -x /etc/rc.d/init.d/$I ]; then

        /etc/rc.d/init.d/$I restart  

      else

        echo "$I is not a valid service..."

      fi

    done

  fi

}

StopService() { 

  if [ ${#@} -gt 0 ]; then  

    for in [email protected]; do

      if [ -x /etc/rc.d/init.d/$I ]; then

        /etc/rc.d/init.d/$I stop  

      else

        echo "$I is not a valid service..."

      fi

    done

  fi

}

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  

}

# Main Function  

ParseOptions [email protected]  

[ $? -ne 0 ] && Usage && exit 5

[ $helpflag -eq 1 ] && Usage && exit 0

if [ $addressflag -ne 1 -o $notifyflag -ne 1 ]; then

  Usage  

  exit 2  

fi

mode=${mode:-mb}

case $notify in

‘master‘)  

  if [ $serviceflag -eq 1 ]; then

      RestartService ${services[*]}  

  fi

  Notify master  

  ;;  

‘backup‘)  

  if [ $serviceflag -eq 1 ]; then

    if "$mode" == ‘mb‘ ]; then

      StopService ${services[*]}  

    else

      RestartService ${services[*]}  

    fi

  fi

  Notify backup  

  ;;  

‘fault‘)  

  Notify fault  

  ;;  

*)  

  Usage  

  exit 4  

  ;;  

esac

(5).给脚本增加执行权限


1

[[email protected] keepalived]# chmod +x notify.sh

(6).将master上脚本复制到slave上


1

[[email protected] keepalived]# scp -p notify.sh [email protected]:/etc/keepalived/

(7).测试一下脚本


1

2

3

4

5

6

7

[[email protected] keepalived]# ./notify.sh -h 

Usage: notify.sh [-m|--mode {mm|mb}] [-s|--service SERVICE1,...] <-a|--address VIP>  <-n|--notify {master|backup|falut}>  

Usage: notify.sh -h|--help  

[[email protected] keepalived]# ./notify.sh --help  

Usage: notify.sh [-m|--mode {mm|mb}] [-s|--service SERVICE1,...] <-a|--address VIP>  <-n|--notify {master|backup|falut}>  

Usage: notify.sh -h|--help  

[[email protected] keepalived]# ./notify.sh -m mb -a 1.1.1.1 -n master

(8).查看一下邮件

注,大家可以看到成功收到邮件,测试成功。在模拟故障时先重启一下keepalived服务。

(9).模拟故障


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

[[email protected] keepalived]# ip addr show #查看一下VIP 

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 00:0c:29:4b:a1:85 brd ff:ff:ff:ff:ff:ff  

    inet 192.168.18.208/24 brd 192.168.18.255 scope global eth0  

    inet 192.168.18.200/32 scope global eth0  

    inet6 fe80::20c:29ff:fe4b:a185/64 scope link   

       valid_lft forever preferred_lft forever  

[[email protected] keepalived]# touch down #进入维护模式  

[[email protected] keepalived]# ll  

总用量 8  

-rw-r--r-- 1 root root    0 8月  22 14:39 down  

-rw-r--r-- 1 root root 1543 8月  22 14:04 keepalived.conf  

-rwxr-xr-x 1 root root 2516 8月  22 14:15 notify.sh  

[[email protected] keepalived]# ip addr show #再次查看VIP  

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 00:0c:29:4b:a1:85 brd ff:ff:ff:ff:ff:ff  

    inet 192.168.18.208/24 brd 192.168.18.255 scope global eth0  

    inet6 fe80::20c:29ff:fe4b:a185/64 scope link   

       valid_lft forever preferred_lft forever

[[email protected] keepalived]# ip addr show #大家可以看到VIP成功移动到slave上 

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 00:0c:29:f9:e6:26 brd ff:ff:ff:ff:ff:ff  

    inet 192.168.18.207/24 brd 192.168.18.255 scope global eth0  

    inet 192.168.18.200/32 scope global eth0  

    inet6 fe80::20c:29ff:fef9:e626/64 scope link   

       valid_lft forever preferred_lft forever

(10).查看一下邮件

注,大家可以看到,主备切换时,会发送邮件报警,好了到这里所有演示全部完成。希望大家有所收获^_^……

时间: 2024-10-08 19:34:54

Linux 高可用(HA)集群之keepalived详解的相关文章

HA集群之keepalived详解/基于keepalived+LVS-DR构建HA主备模型(一)

一.理论部分:     keepalived是vrrp协议的实现:原生设计目的为高可用ipvs服务:keepalived能够配置文件中的定义生成ipvs规则:并能够对各RealServer的健康状态进行检测:  vrrp协议:虚拟冗余路由协议:早期只是主要在路由器上提供的一种非常简单的完成将多个物理设备组建成一个虚拟设备,并且在多个物理设备之间漂移地址一种协议:非常轻量化,性能非常好.而keepalived无非就是通过vrrp协议在Linux主机上通过一个守护进程,把Linux主机扮演成路由器,

HA 集群软件 keepalived 详解3

在前两篇文章中我们队keepalived做了相关介绍,这里环境还是和之前保持的一样,这次我们主要介绍vrrp_script模块,在上次介绍keepalived基础HA功能时用到了vrrp_script这个模块,此模块专门用于对集群中服务自己进行监控,与此模块一起使用的还有track_script模块,在此模块中可以引用监控脚本,命令组合,shell语句等,以实现对服务端口多方面的监控,track_script模块主要用来调用vrrp_script模块使keepalived执行对集群服务资源的检查

高可用(HA)集群原理概述

一.高可用集群(High Availability Cluster) 集群(cluster)就是一组计算机,它们作为一个整体向用户提供一组网络资源.每一个单个的计算机系统都叫集群节点(node).随着业务的增长,集群通过添加新的节点,满足资源的高可扩展性. 计算机硬件和软件易错性不可避免,这样在节点上的服务会不可避免的中断.高可用集群的出现是为保证即使节点失效,而服务能不中断. 高可用集群在一组计算机中,采用主备模式,主节点提供服务,备节点等待:一旦,主节点失效,备节点无需人工的无缝取代主节点提

corosync+pacemaker实现高可用(HA)集群

corosync+pacemaker实现高可用(HA)集群(一) ????重要概念 在准备部署HA集群前,需要对其涉及的大量的概念有一个初步的了解,这样在实际部署配置时,才不至于不知所云 资源.服务与主机(又称节点)的关系: 资源包括vip,httpd,filesystem等: 可整合多个资源形成一个服务: 服务必运行在某个主机上,主机上也可不运行服务(此为空闲主机): 服务里的所有资源应该同时运行在同一个节点上,实现方式有2种: 资源组: 排列约束 资源类型 primitive(或native

iscsI服务的配置+lvm逻辑卷 高可用(HA)集群铺设的业务Apache+iscsI服务的配置

主机环境 redhat6.5 64位 实验环境 服务端1 ip 172.25.29.1 主机名:server1.example.com ricci iscsi  apache     服务端2 ip 172.25.29.2  主机名:server2.example.com ricci iscsi  apache     管理端1 ip 172.25.29.3  主机名:server3.example.com  luci  scsi     管理端2 ip 172.25.29.250     fe

VMware8实现高可用(HA)集群

陈科肇 =========== 操作系统:中标麒麟高级操作系统V6 x86-64 实现软件:中标麒麟高可用集群软件 ======================== 1.环境的规划与配置 硬件要求 服务器 服务器至少需要 2 台,每台服务器至少需要 2 块网卡以做心跳与连接公网使用 存储环境 建议使用一台 SAN/NAS/ISCSI 存储作为数据共享存储空间 软件要求 1.需要在每台服务器上安装中标麒麟 Linux 服务器操作系统 V5.4 以上版本 2.需要将中标麒麟高可用集群软件分别安装在每

Linux集群系列——keepalived详解

keepalived :ipvs -->HAipvs -->VIPvrrp :1master,n backup路由高可用协议keepalived+nginxkeepalived+harproxykeepalived是严重依赖于脚本实现高可用的一种方式 实验实例 :ipvshealth check fall_back 原文地址:https://blog.51cto.com/14087650/2431456

CentOS Linux 负载均衡高可用WEB集群之LVS+Keepalived配置

CentOS Linux 负载均衡高可用WEB集群之LVS+Keepalived配置 LB集群是locd balance集群的简称.翻译成中文是:负载均衡集群的意思:集群是一组相互独立的.通过高速网络互联的计算机相互之间构成一个组合,并以单一的系统的模式加以管理.LVS是Linux Virtual Server的简写,翻译中文是Linux虚拟服务器,是一个虚拟的服务器集群系统. 负载均衡集群:是为了企业提供更为实用,性价比更高的系统机构解决方案.负载均衡集群把用户的请求尽可能的平均分发到集群的各

Linux HA Cluster高可用服务器集群,所谓的高可用不是主机的高可用,而是服务的高可用。

什么叫高可用:一个服务器down掉的可能性多种多样,任何一个可能坏了都有可能带来风险,而服务器离线通常带来的代价是很大的,尤其是web站点,所以当某一台提供服务的的服务器down掉不至于服务终止的就叫高可用. 什么叫心跳:就是将多台服务器用网络连接起来,而后每一台服务器都不停的将自己依然在线的信息很简短很小的通告给同一个网络中的备用服务器的主机,告诉其实主机自己依然在线,其它服务器收到这个心跳信息就认为本机是在线的,尤其是主服务器. 心跳信息怎么发送,由谁来收,其实就是进程中的通信两台主机是没法