暂无图片
暂无图片
暂无图片
暂无图片
暂无图片
首页 / Keepalived实现双机热备

Keepalived实现双机热备

MrJacky博客 2019-08-05
434

Keepalived作用是？？

keepalived的作用是检测后端TCP服务的状态，如果有一台提供TCP服务的后端节点死机，或者工作出现故障，keepalived会及时检测到，并将有故障的节点从系统中剔除，当提供TCP服务的节点恢复并且正常提供服务后keepalived会自动将TCP服务的节点加入到集群中。这些工作都是keepalived自动完成，不需要人工干涉，需要人工做的只是修复发生故障的服务器。

前提


为了测试能够顺利进行，需要先关闭selinux和Firewalld
测试环境信息如下：


keepalived主机：20.20.20.121
keepalived备机：20.20.20.122
web服务器1:     20.20.20.126
web服务器2：20.20.20.127
VIP:           20.20.20.128

一、网络拓扑图

二、部署web服务器

1、两台服务器安装httpd

yum install -y httpd

2、添加首页

web服务器1设置主页
echo "20.20.20.126" >/var/www/html/index.html
web服务器2设置主页
echo "20.20.20.127" >/var/www/html/index.html

3、启动web服务并设置开机自启动

[root@node126 ~]# systemctl start httpd
[root@node126 ~]# systemctl enable httpd
Created symlink from etc/systemd/system/multi-user.target.wants/httpd.service to usr/lib/systemd/system/httpd.service.
[root@node126 ~]#
[root@node127 ~]# systemctl enable httpd
Created symlink from etc/systemd/system/multi-user.target.wants/httpd.service to usr/lib/systemd/system/httpd.service.
[root@node127 ~]#

三、部署keepalived主机服务

1、两台keepalived服务器安装keepalived

yum install openssl openssl-devel keepalived -y

2、keepalived主机配置

keepalived节点1配置文件

[root@node121 ~]# vim etc/keepalived/keepalived.conf

! Configuration File for keepalived

global_defs {
   notification_email {     #指定当keepalived出现问题时，发送邮件给哪些用户
     zhangzhikai@vclusters.com
   }
   notification_email_from jacky@itcloudblog.cn   #指定发件人
   smtp_server localhost    #指定SMTP服务器地址
   smtp_connect_timeout 30   #指定SMTP连接超时时间
   router_id LVS_DEVEL      #标识当前节点名称，不允许重复
}

vrrp_instance VI_1 {    #定义一个实例，一个实例就是一个集群，当前实例名称VI_1可以更改
    state MASTER     #指定该节点为主节点
    interface eth0    #指定VIP的网络接口
    virtual_router_id 51   #指定VRRP组名，主节点和备用节点需要配置相同VRRP组名
    priority 100    #优先级，范围1-254 数字数值比较，越大优先级别越高，主节点优先级必须高于备用节点。
    advert_int 1   #组播信息发送间隔，单位秒，主节点备用节点必须设置一致
    authentication {   #设置验证信息，主节点备用节点必须设置一致
        auth_type PASS   #指定认证方法，PASS简单密码认证
        auth_pass 1111   指定认证所使用的密码，最多8位
    }
    virtual_ipaddress {  #指定VIP，主节点备用节点必须一致
        20.20.20.128
    }
}

virtual_server 20.20.20.128 80 {  对VIP为20.20.20.128，端口号为80的服务器添加相关信息
    delay_loop 6   #keepalived多长时间监测一次真实服务器，单位秒
    lb_algo rr   #LVS调度算法
    lb_kind DR   #LVS-DR模式
#    persistence_timeout 50      同一个IP50秒内的请求都会发到同一个真实服务器，会影响rr算法调度，测试时可以注释掉
    protocol TCP    #4层协议

    real_server 20.20.20.126 80 {  #对IP为20.20.20.126，端口号为80的真实服务器添加相关信息
        weight 1  #指定权重，默认为1
        TCP_CHECK {
            connect_timeout 3     #连接超时时间，默认5秒
            nb_get_retry 3       #重试次数，默认1次
            delay_before_retry 3    #重试时间间隔，默认1秒
            connect_port 80    #监测端口号
            }
        }
    real_server 20.20.20.127 80 {    　#对IP为20.20.20.127，端口号为80的真实服务器添加相关信息
        weight 1
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
}

3、keepalived备服务器的keepalived.conf的配置

[root@node122 ~]# vim etc/keepalived/keepalived.conf

! Configuration File for keepalived

global_defs {
   notification_email {
     zhangzhikai@vclusters.com
   }
   notification_email_from jacky@itcloudblog.cn
   smtp_server localhost
   smtp_connect_timeout 30
   router_id LVS_BACKUP     ###修改为LVS_BACKUP
}

vrrp_instance VI_1 {
    state BACKUP         ###修改为BACKUP
    interface eth0
    virtual_router_id 51
    priority 90         ###备用修改为90 主节点为100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        20.20.20.128
    }
}

virtual_server 20.20.20.128 80 {
    delay_loop 6
    lb_algo rr
    lb_kind DR
#    persistence_timeout 50
    protocol TCP

    real_server 20.20.20.126 80 {
        weight 1
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
            }
        }
    real_server 20.20.20.127 80 {
        weight 1
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
            connect_port 80
        }
    }
}

/etc/keepalived/keepalived.conf为keeplived的主配置文件。以上配置state表示主节点为：20.20.20.121，副节点为：20.20.20.122。虚拟为IP：20.20.20.128，后端的真实服务器为：20.20.20.126和20.20.20.127，当通过20.20.20.128访问web服务器时，自动转到后端真实服务器，后端节点的权重相同，类似轮询模式。

四、keepalived启动服务与测试

1、启动keepalived服务

systemctl start keepalived
systemctl enable keepalived

2、测试VIP漂移
查看主节点和备用节点的ip地址

[root@node121 ~]# ip add
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
    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 1e:c0:57:cd:d5:03 brd ff:ff:ff:ff:ff:ff
    inet 20.20.20.121/24 brd 20.20.20.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet 20.20.20.128/32 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::1cc0:57ff:fecd:d503/64 scope link 
       valid_lft forever preferred_lft forever
[root@node122 ~]# ip add
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
    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 8e:39:b9:7e:1b:a4 brd ff:ff:ff:ff:ff:ff
    inet 20.20.20.122/24 brd 20.20.20.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::8c39:b9ff:fe7e:1ba4/64 scope link 
       valid_lft forever preferred_lft forever
[root@node122 ~]#

停止主节点的keepalived，再查看主节点和备用节点IP

[root@node121 ~]# systemctl stop keepalived
[root@node121 ~]# ip add
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
    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 1e:c0:57:cd:d5:03 brd ff:ff:ff:ff:ff:ff
    inet 20.20.20.121/24 brd 20.20.20.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::1cc0:57ff:fecd:d503/64 scope link 
       valid_lft forever preferred_lft forever
[root@node121 ~]#
[root@node122 ~]# ip add
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
    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 8e:39:b9:7e:1b:a4 brd ff:ff:ff:ff:ff:ff
    inet 20.20.20.122/24 brd 20.20.20.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet 20.20.20.128/32 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::8c39:b9ff:fe7e:1ba4/64 scope link 
       valid_lft forever preferred_lft forever
[root@node122 ~]#

测试通过需要再次开启keepalived服务。

五、配置真实服务器回环接口

1、配置node126web服务器回环口lo:1为VIP

[root@node126 ~]# cd etc/sysconfig/network-scripts/
[root@node126 ~]# cd etc/sysconfig/network-scripts/
[root@node126 network-scripts]# cp ifcfg-lo ifcfg-lo:1
DEVICE=lo:1
IPADDR=20.20.20.128
NETMASK=255.255.255.255
#NETWORK=127.0.0.0
# If you're having problems with gated making 127.0.0.0/8 a martian,
# you can change this to something else (255.255.255.255, for example)
#BROADCAST=127.255.255.255
ONBOOT=yes
NAME=loopback
[root@node126 ~]# systemctl restart network

2、配置node12web服务器回环口lo:1为VIP

[root@node127 ~]# cd etc/sysconfig/network-scripts/
[root@node127 ~]# cd etc/sysconfig/network-scripts/
[root@node127 network-scripts]# cp ifcfg-lo ifcfg-lo:1
DEVICE=lo:1
IPADDR=20.20.20.128
NETMASK=255.255.255.255
#NETWORK=127.0.0.0
# If you're having problems with gated making 127.0.0.0/8 a martian,
# you can change this to something else (255.255.255.255, for example)
#BROADCAST=127.255.255.255
ONBOOT=yes
NAME=loopback
[root@node127 ~]# systemctl restart network

六、测试

1、客户端访问VIP页面轮训到多台web服务器


2、模拟web1服务器故障

关闭web节点之前查看有两台web服务器

[root@node121 ~]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  20.20.20.128:80 rr
  -> 20.20.20.126:80              Route   1      0          0
  -> 20.20.20.127:80              Route   1      0          0
[root@node121 ~]#

关闭web1服务器

[root@node126 ~]# systemctl stop httpd

关闭web节点之后，将故障web服务器踢除操作

[root@node121 ~]# ipvsadm -L -n
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  20.20.20.128:80 rr
  -> 20.20.20.127:80              Route   1      0          0         
[root@node121 ~]#

七、实现keepalived两台服务器互为主从

在之前基础配置上实现keepalived互为主从，添加另外一个VIP:20.20.20.131，本次node122作为master，node121作为backup。

1、配置keepalived主机


（1）、在node121修改配置文件，添加实例和对应主机

vrrp_instance VI_2 {
    state BACKUP
    interface eth0
    virtual_router_id 52
    priority 90
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 2222
    }
    virtual_ipaddress {
        20.20.20.131
    }
}

virtual_server 20.20.20.131 80 {
    delay_loop 6
    lb_algo rr
    lb_kind DR
#    persistence_timeout 50
    protocol TCP

    real_server 20.20.20.126 80 {
        weight 1
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        connect_port 80
            }
    }
    real_server 20.20.20.127 80 {
        weight 1
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        connect_port 80
        }
    }
}

重启keepalived服务

[root@node121 ~]# systemctl restart keepalived

（2）、在node122修改配置文件，添加实例和对应主机

vrrp_instance VI_2 {
    state MASTER
    interface eth0
    virtual_router_id 52
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 2222
    }
    virtual_ipaddress {
        20.20.20.131
    }
}

virtual_server 20.20.20.131 80 {
    delay_loop 6
    lb_algo rr
    lb_kind DR
#    persistence_timeout 50
    protocol TCP

    real_server 20.20.20.126 80 {
        weight 1
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        connect_port 80
            }
    }
    real_server 20.20.20.127 80 {
        weight 1
        TCP_CHECK {
            connect_timeout 3
            nb_get_retry 3
            delay_before_retry 3
        connect_port 80
        }
    }
}

重启keepalived服务

[root@node122 ~]# systemctl restart keepalived


2、配置真实web服务器回环口


（1）、在node121配置回环口

[root@node126 ~]# cd etc/sysconfig/network-scripts/
[root@node126 network-scripts]# cp ifcfg-lo:1 ifcfg-lo:2
DEVICE=lo:2
IPADDR=20.20.20.131
NETMASK=255.255.255.255
#NETWORK=127.0.0.0
# If you're having problems with gated making 127.0.0.0/8 a martian,
# you can change this to something else (255.255.255.255, for example)
#BROADCAST=127.255.255.255
ONBOOT=yes
NAME=loopback

重启网络服务

（2）、在node122配置回环口

DEVICE=lo:2
IPADDR=20.20.20.131
NETMASK=255.255.255.255
#NETWORK=127.0.0.0
# If you're having problems with gated making 127.0.0.0/8 a martian,
# you can change this to something else (255.255.255.255, for example)
#BROADCAST=127.255.255.255
ONBOOT=yes
NAME=loopback

重启网络服务

[root@node127 network-scripts]# systemctl restart network


3、查看keepalived节点上VIP信息


（1）、在node121节点查看VIP信息

[root@node121 ~]# ip a sh dev eth0    ##node121节点上只有128，没有131 vip地址
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 1e:c0:57:cd:d5:03 brd ff:ff:ff:ff:ff:ff
    inet 20.20.20.121/24 brd 20.20.20.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet 20.20.20.128/32 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::1cc0:57ff:fecd:d503/64 scope link 
       valid_lft forever preferred_lft forever
[root@node121 ~]#

（2）、在node122节点查看VIP信息

[root@node122 ~]# ip a sh dev eth0   ##node122节点上只有131，没有128 vip地址
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 8e:39:b9:7e:1b:a4 brd ff:ff:ff:ff:ff:ff
    inet 20.20.20.122/24 brd 20.20.20.255 scope global eth0
       valid_lft forever preferred_lft forever
    inet 20.20.20.131/32 scope global eth0
       valid_lft forever preferred_lft forever
    inet6 fe80::8c39:b9ff:fe7e:1ba4/64 scope link 
       valid_lft forever preferred_lft forever
[root@node122 ~]#


4、验证测试


keepalived部署过程中说明：
（1）、在修改keepalived配置文件中注意{}对称
（2）、主节点和备用节点 route_id 集群名称等 
（3）、在测试过程中不轮训的情况，主要原因是：keepalived参数和lvs参数的问题

keepalived.conf配置文件注释persistence_timeout 50参数
查看ipvsadm -L --timeout
默认是Timeout (tcp tcpfin udp): 900 120 300 修改为：Timeout (tcp tcpfin udp): 1 2 1
修改命令：ipvsadm --set 1 2 1

原因解释：
检查keepalived软件配置，发现virtual_server配置了persistence_timeout 连接保持，意思就是在这个一定时间内会讲来自同一用户（根据ip来判断的）访问到同一个real server。
配置连接保持后用户访问到哪个后端由persistence_timeout和lvs的超时共同决定。


数据库
文章转载自MrJacky博客，如果涉嫌侵权，请发送邮件至：contact@modb.pro进行举报，并提供相关证据，一经查实，墨天轮将立刻删除相关内容。

评论