标签:高可用 集群 keepalived
一、承接上一篇文章,按上面拓扑做修改:
1.新增一台Director,eth1的IP 192.168.1.5
2.旧Director添加一起网卡eth0(IP 10.0.0.1/24)直连新Director的eth0(IP 10.0.0.2/24)
3.删除.旧Director的eth1:0
二、两台Director都进行以下操作:
1.关闭ipvsadm:
chkconfig ipvsadm off
service ipvsadm off
2..通过ssh-keygen互信
ssh-keygen -t rsa -f ~/.ssh/idrsa -P ‘‘
ssh-copy-id -i .ssh/idrsa.pub root@192.168.1.5
3..安装keepalived
yum -i install keepalvied
4.设置配置文件
注意:
(1)红框是在backup的director上要进行修改的。
(2)所有大括号前必须有空格,否则会有异常,且不好排查。
vim /etc/keepalived/keepalived.conf
5.创建状态变化运行的脚本(可根据实际需要创建,这里只提供一个测试例子)
vim /etc/keepalived/status.sh
#!/bin/bash
echo "`date ‘+%F %T’`$1 change status to $2 " >>/etc/keepalived/status.txt
chmod +x /etc/keepalived/stauts.sh
6.启动服务
service keepalived start
chkconfig keepalived on
一开始旧Director为master
实验:
在旧Director上创建文件(touch /etc/keepalived/down)则会变成backup(优先级由原来的100变成了90),并且系统自动运行脚本 /etc/keepalived/status.sh 192.168.1.1 backup ,自动将状态改变写进status.txt文件里,就可通过tail -f /etc/keepalived/status.txt 查看当前keepalived的模式状态
rm -f /etc/keepalived/down ,则旧Director又变成master
将旧Director的eth1断开,则旧Director又变成master,而新Director变成fault
在httpd服务器上将主页移走:mv index.html index.html.bak ,Director的上ipvsadm可以看到将此httpd服务器移除。
本文出自 “辛-碌-力-成” 博客,请务必保留此出处http://orzorz.blog.51cto.com/4228156/1702201
高可用集群实验二:keepalived+lvs实现故障自迁移
标签:高可用 集群 keepalived
原文地址:http://orzorz.blog.51cto.com/4228156/1702201