Linux 高可用keepalive

Posted

tags:

篇首语:本文由小常识网(cha138.com)小编为大家整理,主要介绍了Linux 高可用keepalive相关的知识,希望对你有一定的参考价值。

技术分享图片
vrrp:虚拟路由冗余协议(Virtual Router Redundancy Protocol),冗余路由单点故障,实现把一个地址绑定到n台服务器上(只生效一台),通过组播传递信号,当一个点故障时候,选举另外点来当路由。
keepalive工作流程:监视进程watchDog监测 vrrp跟checkers,监测过程中触发系统调用跟邮件服务,来调整浮动ip,内置ipvs规则装饰跟启动ipvs脚本。

安装

网卡组播必须开启

ens34: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500

ip link set multicast on  dev ens33
ip link set multicast off dev ens33

安装

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

vrrp使用

主节点配置

[[email protected] keepalived]# vim /etc/keepalived/keepalived.conf
global_defs {
   notification_email {
        [email protected]
   }
   notification_email_from [email protected]
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id node1                   #单个id
   vrrp_mcast_group4 224.1.101.33    #组播地址
}

vrrp_instance VI_1 {
    state MASTER                  #当前节点在此虚拟路由器上的初始状态;只能有一个是MASTER,余下的都应该为BACKUP
    interface ens34               #通告进行所需 接口
    virtual_router_id 51          #虚拟路由器标识:VRID(0-255),唯一标识虚拟路由器
    priority 100                  #优先级
    advert_int 1                  #vrrp通告的时间间隔,默认1s
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        192.168.1.199 
    }
    #状态改变时候出发钩子函数
    notify_master "/etc/keepalived/notify.sh master"  #当前节点成为主节点时触发的脚本
    notify_backup "/etc/keepalived/notify.sh backup"  #当前节点转为备节点时触发的脚本
    notify_fault "/etc/keepalived/notify.sh fault"   #当前节点转为“失败”状态时触发的脚本
}

从节点(区别主节点有3个地方)

global_defs {
   notification_email {
        [email protected]
   }
   notification_email_from [email protected]
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id node2                         #1.标识
   vrrp_mcast_group4 224.1.101.33
}

vrrp_instance VI_1 {
    state BACKUP                           #2.备用
    interface ens34
    virtual_router_id 51
    priority 96                            #3.优先级
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        192.168.1.199 
    }
    notify_master "/etc/keepalived/notify.sh master"
    notify_backup "/etc/keepalived/notify.sh backup"
    notify_fault "/etc/keepalived/notify.sh fault"
}

邮件通知脚本(如果是双主,脚本后面加参数区分)

[[email protected] keepalived]# vim notify.sh
#!/root/.virtualenvs/shellenv/bin/python

from raven import Client
import subprocess

import sys

if len(sys.argv) < 2:
    exit()

notify = sys.argv[1]

interface = ‘ens34‘
res = subprocess.getstatusoutput("ip addr show %s|egrep -o ‘inet [0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}‘|egrep -o ‘[^a-z ].*‘"%(interface))
ip = res[1]
ip = ip.replace(‘\n‘, ‘ ‘)

res = subprocess.getstatusoutput("hostname")
hostname = res[1]

res = subprocess.getstatusoutput("uptime")
cpu = res[1]

res = subprocess.getstatusoutput("free -h")
memory = res[1]

res = subprocess.getstatusoutput("df -h")
disk = res[1]

notify_message = {‘master‘:‘Change to master‘, ‘backup‘:‘Change to backup‘, ‘fault‘:‘Server error‘}
message = notify_message.get(notify,‘{%s} no this notify‘%(notify))

info = ‘‘‘
-------- info of %s  -----
Hostname:%s

Interface:%s

IP:%s

CPU:%s

Memory:
%s

Disk:
%s

Message: %s
‘‘‘ % (hostname, hostname, interface, ip, cpu, memory,disk,  message)

client = Client(‘https://[email protected]/1219450‘)
client.captureMessage(info)

启动

[[email protected] keepalived]# systemctl  start keepalived

主节点通告

[[email protected] keepalived]# tcpdump -i ens34 -nn host 224.1.101.33
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on ens34, link-type EN10MB (Ethernet), capture size 262144 bytes
13:04:17.240395 IP 192.168.1.200 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20

备用节点收通告

[[email protected] keepalived]# tcpdump -i ens34 -nn host 224.1.101.33
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on ens34, link-type EN10MB (Ethernet), capture size 262144 bytes
13:05:36.437749 IP 192.168.1.200 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20

主节点查看

[[email protected] keepalived]# ip a
3: ens34: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
    link/ether 00:0c:29:46:a6:40 brd ff:ff:ff:ff:ff:ff
    inet 192.168.1.200/24 brd 192.168.1.255 scope global ens34
       valid_lft forever preferred_lft forever
    inet 192.168.1.198/32 scope global ens34:0
       valid_lft forever preferred_lft forever
    inet6 fe80::20c:29ff:fe46:a640/64 scope link
       valid_lft forever preferred_lft forever

从节点

此时从节点没有 ens34:0

主节点故障

#ip到从节点上
[[email protected] keepalived]# ifconfig
ens34:0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 192.168.1.199  netmask 255.255.255.255  broadcast 0.0.0.0
        ether 00:0c:29:34:f5:84  txqueuelen 1000  (Ethernet)

主节点恢复,因为优先级比从节点高,所以ip又到主节点上

邮箱读取
主服务开启:
技术分享图片
技术分享图片

主节点停掉服务
技术分享图片

日志配置

[[email protected] keepalived]# vim /etc/sysconfig/keepalived
KEEPALIVED_OPTIONS="-D -S 3"

[[email protected] keepalived]# vim /etc/rsyslog.conf
local3.*                                                /var/log/keepalive.log

[[email protected] keepalived]# systemctl restart rsyslog
[[email protected] keepalived]# systemctl restart keepalived

lvs双主模型

双主配置单节点配置,另外节点只需改动主从、routerid、优先级即可

global_defs {
   notification_email {
        [email protected]
   }
   notification_email_from [email protected]
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id node1
   vrrp_mcast_group4 224.1.101.33
}

vrrp_instance VI_1 {
    state MASTER
    interface ens34
    virtual_router_id 51
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    virtual_ipaddress {
        192.168.1.198 dev ens34 label ens34:0
    }

    notify_master "/etc/keepalived/notify.sh master"
    notify_backup "/etc/keepalived/notify.sh backup"
    notify_fault "/etc/keepalived/notify.sh fault"
}

vrrp_instance VI_2 {
    state BACKUP
    interface ens34
    virtual_router_id 52
    priority 98
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 2222
    }
    virtual_ipaddress {
        192.168.1.197
    }

    notify_master "/etc/keepalived/notify.sh master"
    notify_backup "/etc/keepalived/notify.sh backup"
    notify_fault "/etc/keepalived/notify.sh fault"
}

virtual_server 192.168.1.198 80 {
    delay_loop 1                     
    lb_algo wrr                   
    lb_kind DR                   
    protocol TCP               
    sorry_server 192.168.1.200 80        

    real_server 192.168.1.202 80{
        weight 1
        HTTP_GET {
                url {
                        path /index.html
                        status_code 200
                }
                nb_get_retry 3
                delay_before_retry 2
                connect_timeout 3
        }
    }
    real_server 192.168.1.203 80{
        weight 1
        TCP_CHECK {
                nb_get_retry 3
                delay_before_retry 2
                connect_timeout 3
        }
    }
}

virtual_server 192.168.1.197 80 {
    delay_loop 1
    lb_algo wrr
    lb_kind DR
    protocol TCP
    sorry_server 192.168.1.200 80

    real_server 192.168.1.202 80{
        weight 1
        HTTP_GET {
                url {
                        path /index.html
                        status_code 200
                }
                nb_get_retry 3
                delay_before_retry 2
                connect_timeout 3
        }
    }
    real_server 192.168.1.203 80{
        weight 1
        TCP_CHECK {
                nb_get_retry 3
                delay_before_retry 2
                connect_timeout 3
        }
    }
}
[[email protected] keepalived]# ipvsadm -ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.1.197:80 wrr
  -> 192.168.1.202:80             Route   1      0          0
  -> 192.168.1.203:80             Route   1      0          0
TCP  192.168.1.198:80 wrr
  -> 192.168.1.202:80             Route   1      0          0
  -> 192.168.1.203:80             Route   1      0          0
[[email protected] keepalived]# ipvsadm -ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  192.168.1.197:80 wrr
  -> 192.168.1.202:80             Route   1      0          0
  -> 192.168.1.203:80             Route   1      0          0
TCP  192.168.1.198:80 wrr
  -> 192.168.1.202:80             Route   1      0          0
  -> 192.168.1.203:80             Route   1      0          0

由此可见:ipvs规则两台机器配置是一样的,只是vip是飘动的,keepalive针对lvs还实现了健康状态监测

marvindeMacBook-Pro:~ marvin$ curl http://192.168.1.197/
node4
marvindeMacBook-Pro:~ marvin$ curl http://192.168.1.197/
node3

#停掉node3节点
marvindeMacBook-Pro:~ marvin$ curl http://192.168.1.197/
node4
marvindeMacBook-Pro:~ marvin$ curl http://192.168.1.197/
node4

自定义监测模型-nginx

配置手动下线文件脚本

[[email protected] keepalived]# vim down.sh
 [[ -f /etc/keepalived/down ]] && exit 1 || exit 0

配置文件

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

 global_defs {
   notification_email {
        [email protected]
   }
   notification_email_from [email protected]
   smtp_server 127.0.0.1
   smtp_connect_timeout 30
   router_id node1
   vrrp_mcast_group4 224.1.101.33
}

vrrp_script chk_down {
        script "/etc/keepalived/down.sh"
        interval 1
        weight -20
        fall 1
        rise 1
}

vrrp_script chk_nginx {
    script "pidof nginx && exit 0 || exit 1"
    interval 1
    weight -3
    fall 2
    rise 2      #有可能重启后又down 所以2次比较合理
}

vrrp_instance VI_1 {
    state MASTER
    interface ens34
    virtual_router_id 51
    priority 100
    advert_int 1
    authentication {
        auth_type PASS
        auth_pass 1111
    }
    track_script {
        chk_down
        chk_nginx
    }
    virtual_ipaddress {
        192.168.1.198/24
    }

    notify_master "/etc/keepalived/notify.sh master"
    notify_backup "/etc/keepalived/notify.sh backup"
    notify_fault "/etc/keepalived/notify.sh fault"
}

/etc/keepalived/notify.sh主要在里面加入

res = subprocess.getstatusoutput("systemctl start mynginx")
res = subprocess.getstatusoutput("pidof nginx")
nginxpid = res[1]

测试

[[email protected] keepalived]# tcpdump -i ens34 -nn host 224.1.101.33

[[email protected] keepalived]# touch down
15:29:27.073406 IP 192.168.1.200 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
15:29:28.074393 IP 192.168.1.200 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 80, authtype simple, intvl 1s, length 20
15:29:28.074710 IP 192.168.1.201 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 99, authtype simple, intvl 1s, length 20

[[email protected] keepalived]#  rm -f down
15:30:17.179289 IP 192.168.1.201 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 99, authtype simple, intvl 1s, length 20
15:30:17.180088 IP 192.168.1.200 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20

[[email protected] keepalived]# systemctl stop mynginx
15:31:01.251469 IP 192.168.1.200 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
15:31:02.253183 IP 192.168.1.200 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
15:31:03.254806 IP 192.168.1.200 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 97, authtype simple, intvl 1s, length 20
15:31:03.255137 IP 192.168.1.201 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 99, authtype simple, intvl 1s, length 20
15:30:17.180088 IP 192.168.1.200 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s, length 20
15:30:18.181628 IP 192.168.1.200 > 224.1.101.33: VRRPv2, Advertisement, vrid 51, prio 100, authtype simple, intvl 1s,

随后自动启动 是因为    notify_backup "/etc/keepalived/notify.sh backup"
这个脚本写了nginx 启动

邮件查看
技术分享图片
技术分享图片

以上是关于Linux 高可用keepalive的主要内容,如果未能解决你的问题,请参考以下文章

Keepalive配置文件说明和实现高可用

linux之LVS的DR模式搭建以及配置高可用管理LVS(heartbeat,keepalive)

LinuxCentOS系统Nginx+Keepalive主从自动切换高可用

Linux集群之keepalive+Nginx

harbor基于keepalive高可用部署

LVS+Keepalive的高可用群集