配置Failover Clustering
1.将各节点服务器加入域(应该是还没加,这两天陆陆续续写自己都有点忘了)
各Node服务为:
HY-A-BL2
HY-A-BL3
HY-A-BL9
HY-A-BL10
2.同时将MS的补丁更新至最新及各Node一致。
3.打开Failover Cluster Management
![](https://s1.51cto.com/attachment/201009/201009241285315118198.png)
4.验证Cluster Failover的配置条件
![](https://s1.51cto.com/attachment/201009/201009241285317610969.png)
![](https://s1.51cto.com/attachment/201009/201009241285317914311.png)
添加各Node服务器,并按Next
![](https://s1.51cto.com/attachment/201009/201009241285318074806.png)
![](https://s1.51cto.com/attachment/201009/201009241285318194065.png)
![](https://s1.51cto.com/attachment/201009/201009251285374114289.png)
验证配置显示符合Failover Cluster的要求,但还是有一些错误,可以查看详细报表,并查看原因
![](https://s1.51cto.com/attachment/201009/201009241285318799827.png)
5.建立Cluster Failover
![](https://s1.51cto.com/attachment/201009/201009241285319088014.png)
![](https://s1.51cto.com/attachment/201009/201009241285319203427.png)
输入Clsuter Name及IP地址,并按两个Next后,建立cluster.
![](https://s1.51cto.com/attachment/201009/201009241285319465529.png)
设定Failover Cluster,下图为建立好的界面,包括4个节点
![](https://s1.51cto.com/attachment/201009/201009241285333568733.png)
为了方便查看,修改网络定义
将192.168.1.0网络重命名为HeartBeat
![](https://s1.51cto.com/attachment/201009/201009241285336518078.png)
并修改网络连接属性,用于各Node间的通信,用于Live Migration。
![](https://s1.51cto.com/attachment/201009/201009241285336621748.png)
将172.20.184.0网络用于连接Share Storage的网络名定义为iSCSI
![](https://s1.51cto.com/attachment/201009/201009241285336688949.png)
对该网络名单击右键“属性”修改,如下图:
![](https://s1.51cto.com/attachment/201009/201009241285336744436.png)
将172.20.179.0网络改为MGT
![](https://s1.51cto.com/attachment/201009/201009241285336794561.png)
并确认网络属性
![](https://s1.51cto.com/attachment/201009/201009241285336917714.png)
修改Storage名,在下图中将Witness盘及CSV盘分改名为如下图:(以下为已添加CSV的界面)
![](https://s1.51cto.com/attachment/201009/201009241285336994555.png)
添加Cluster Shared Volumes,并选中刚定义的CSV盘
![](https://s1.51cto.com/attachment/201009/201009241285337061146.png)
添加后,
![](https://s1.51cto.com/attachment/201009/201009241285337321209.png)
配置Witness盘
![](https://s1.51cto.com/attachment/201009/201009241285337954875.png)
按建议的方式建立
![](https://s1.51cto.com/attachment/201009/201009241285338362904.png)
选择Witness盘,
![](https://s1.51cto.com/attachment/201009/201009241285338660997.png)
![](https://s1.51cto.com/attachment/201009/201009241285338833875.png)
在HYPER-V管理器内建立一台VM,(过程不再介绍)
以下几点设置需要注意,见下图
![](https://s1.51cto.com/attachment/201009/201009241285339850981.png)
自动启动选择“无”
![](https://s1.51cto.com/attachment/201009/201009241285339953454.png)
添加Hyper-V服务,
![](https://s1.51cto.com/attachment/201009/201009241285339150027.png)
![](https://s1.51cto.com/attachment/201009/201009241285340309709.png)
选择虚拟机后,按Next后完成添加。
![](https://s1.51cto.com/attachment/201009/201009241285340425678.png)
此时该台VM就有了高可用性,(这篇文章写在部署后,所以该Cluster上已经有了多台VM)
![](https://s1.51cto.com/attachment/201009/201009241285340774484.png)
我们安装完后可测试Live Migration。我有录下测试时的过程,基本上就丢一个包。请看附件(大小受限,所以不是很清晰)。
Live Migration对我来说非常有用,当要对一台服务器进行维护,如安装更新或是升级硬件时,可不中断业务将VM迁移至另一个节点上。不必再停止业务将VM关机或是停机再做迁移。
转载于:https://blog.51cto.com/sting/396648