最近发现公司云平台服务器的vip有丢失的现象,查看keepalived日志

Jun   :: lb1 dhclient: DHCPREQUEST of 10.0.0.2 on eth0 to 10.0.0.3 port  (xid=0x6deab016)
Jun :: lb1 dhclient: DHCPNAK from 10.0.0.3 (xid=0x6deab016)
Jun :: lb1 dhclient: DHCPDISCOVER on eth0 to 255.255.255.255 port interval (xid=0x37e1db6a)
Jun :: lb1 dhclient: DHCPREQUEST of 10.0.0.2 on eth0 to 255.255.255.255 port (xid=0x37e1db6a)
Jun :: lb1 dhclient: DHCPOFFER of 10.0.0.2 from 10.0.0.3
Jun :: lb1 dhclient: DHCPACK of 10.0.0.2 from 10.0.0.3
Jun :: lb1 dhclient: bound to 10.0.0.2 -- renewal in seconds.
Jun :: lb1 ntpd[]: Deleting interface # eth0, 10.0.255.254#, interface stats: received=, sent=, dropped=, active_time= secs
Jun :: lb1 ntpd[]: peers refreshed

发现是云平台DHCP刷新直接把vip删掉,但是由于俩台机器的keepalived vrrp_script监控应用服务都是正常的,所以主备关系没有改变,如果没有特意加ping vip的策略,会导致keepalived没有发现此问题。

解决方法:

1、如果应平台的服务器可以改成固定IP,可以将网卡改为固定IP,这样就不会使用DHCP了

cat /etc/sysconfig/network-scripts/ifcfg-eth0
DEVICE=eth0
TYPE=Ethernet
ONBOOT=yes
NM_CONTROLLED=no
BOOTPROTO=static
IPADDR=10.0.0.2
NETMASK=255.255.255.0
GATEWAY=10.0.0.1

2、如果不可以更改网卡IP,就在keepalived配置文件中增加dont_track_primary参数

vrrp_instance VI_1 {
state BACKUP
interface eth0
dont_track_primary
virtual_router_id
priority
advert_int
nopreempt
authentication {
auth_type PASS
auth_pass
}
virtual_ipaddress {
10.0.0.17
}
}

dont track primary:忽略VRRP的interface错误

如果有两块网卡,可以搭配track interface使用

track interface:跟踪接口,设置额外的监控,里面任意一块网卡出现问题,都会进入故障(FAULT)状态,例如,用nginx做均衡器的时候,内网必须正常工作,如果内网出问题了,这个均衡器也就无法运作了,所以必须对内外网同时做健康检查

vrrp_instance http {
state MASTER
interface eth0
dont_track_primary
track_interface {
eth0
eth1
}

官方原文

# Ignore VRRP interface faults (default unset)
dont_track_primary # optional, monitor these as well.
# go to FAULT state if any of these go down if unweighted.
# When a weight is specified in track_interface, instead of setting the vrrp
# instance to the FAULT state in case of failure, its priority will be
# increased by the weight when the interface is up (for positive weights),
# or decreased by the weight's absolute value when the interface is down
# (for negative weights). The weight must be comprised between - and +
# inclusive. is the default behaviour which means that a failure implies a
# FAULT state. The common practice is to use positive weights to count a
# limited number of good services so that the server with the highest count
# becomes master. Negative weights are better to count unexpected failures
# among a high number of interfaces, as it will not saturate even with high
# number of interfaces.
track_interface {
eth0
eth1
eth2 weight <-..>
...
} # add a tracking script to the interface
# (<SCRIPT_NAME> is the name of the vrrp_track_script entry)
# The same principle as track_interface can be applied to track_script entries,
# except that an unspecified weight means that the default weight declared in
# the script will be used (which itself defaults to ).

参考

https://serverfault.com/questions/601670/keepalived-vip-removed-with-dhcp-renewal

https://www.keepalived.org/manpage.html

keepalived vip removed with dhcp renewal【原创】的更多相关文章

  1. keepalived的vip无法ping通【原创】

    今天收到redis的keepalived vip无法ping通的告警,查看服务器和服务时发现vip在服务器上,服务也正常.只能在本机ping通,跨网段无法ping通.切换keepalived vip至 ...

  2. keepalived vip 没有生成或者生成了ping不通?

    1 问题现象:keepalived已启动但vip 没有生成./var/log/messages日志不断刷屏 tail /var/log/messages Nov :: cache-redis- Kee ...

  3. OpenStack neutron 环境云主机使用keepalived vip + 给vip绑定浮动IP 步骤及注意事项

    在openstack环境创建的多台云主机配置keepalived作主备,默认情况下无法生效,直接对云主机一张网卡配置两个IP进行测试也是同样结果,因为: 可以看到,port所在的宿主机上iptable ...

  4. keepalived vip漂移基本原理及选举算法

    keepalived可以将多个无状态的单点通过虚拟IP(以下称为VIP)漂移的方式搭建成一个高可用服务,常用组合比如 keepalived+nginx,lvs,haproxy和memcached等.它 ...

  5. (转)小谈keepalived vip漂移原理与VRRP协议

    背景:之前搭建过keepalived双机热备的集群,但对其中的原理不甚理解,看完就忘了,所有有必要深入的学习下. 简介 什么是keepalived呢?keepalived是实现高可用的一种轻量级的技术 ...

  6. 浅析keepalived vip漂移原理与VRRP协议

    2017-01-18 Martin 开源技术社区 简介 什么是keepalived呢?keepalived是实现高可用的一种轻量级的技术手段,主要用来防止单点故障(单点故障是指一旦某一点出现故障就会导 ...

  7. Nginx+Keepalived+VIP漂移实现HA高可用技术之详细教程

    https://www.cnblogs.com/zcc666/p/13141626.html  这个是nginx安装教程地址 https://www.cnblogs.com/zcc666/p/1313 ...

  8. 三主机配置 keepalived VIP高可用

    三台主机: 192.168.33.134 192.168.33.136 192.168.33.137   实验前: 关闭selinux 和iptables     1). 192.168.33.134 ...

  9. Redis主从复制+Keepalived+VIP漂移实现HA高可用技术之详细教程

    1.大家可以先看我的单台Redis安装教程,链接在此点击Redis在CentOS for LInux上安装详细教程 2.第一台redis配置,是正常配置.作为MASTER主服务器,第二台redis的配 ...

随机推荐

  1. dede 对 单个 字段 编辑

    {dede:field.body runphp='yes'} $body = str_replace("src=\"/uploads","src=\" ...

  2. 学习python的日常

    今天是开始正式接触python语言的第一天,然后来自前辈的知道开始了在学习过程当中用博客来记录自己的学习历程,以供自己更快地掌握这门编程语言. 大概的总结一下的话,还是按照我的编程的习惯,要学写代码, ...

  3. 设置grep高亮显示匹配项和基本用法

    设置grep高亮显示匹配项 方法1:设置别名 编辑vim~/.bashrc 添加如下一行内容: alias grep = 'grep --color=auto' source ~/.bashrc // ...

  4. Problem G: STL——整理唱片(list的使用)

    #include<iostream> #include<list> #include<iterator> #include<algorithm> usi ...

  5. python 全局声明 global

    https://www.cnblogs.com/Lin-Yi/p/7305364.html 在基本的python语法当中,一个函数可以随意读取全局数据,但是要修改全局数据的时候有两种方法:1 glob ...

  6. Java精通并发-同步方法访问标志与synchronized关键字之间的关系

    继续基于上一次https://www.cnblogs.com/webor2006/p/11428811.html来研究synchronized关键字在字节码中的表现,在上一次文末提出了一个这样的问题: ...

  7. 微信小程序~App.js中获取用户信息

    (1)代码:主要介绍下获取用户信息部分 onLaunch: function () { // 展示本地存储能力 var logs = wx.getStorageSync('logs') || [] l ...

  8. 均分纸牌(Noip2002)

    1320:[例6.2]均分纸牌(Noip2002) 时间限制: 1000 ms         内存限制: 65536 KB提交数: 3537     通过数: 1839 [题目描述] 有n堆纸牌,编 ...

  9. test20190815 NOIP2019 模拟题

    100+60+40=200,被后面两个题卡着我很不爽. 立方数 [问题描述] 作为 XX 战队的狂热粉丝,MdZzZZ 看到了自己心仪的队伍在半决赛落败,顿时 心灰意冷.看着自己手中的从黄牛那里抢来的 ...

  10. 创建型模式(过渡模式) 简单工厂模式(Simple Factory)

    简单工厂模式(Simple Factory Pattern)属于类的创建型模式,又叫静态工厂方法模式(Static FactoryMethod Pattern) 是通过专门定义一个类来负责创建其他类的 ...