How to deploy OS in a cluster

最新推荐文章于 2023-10-15 12:21:54 发布

zhxue123

最新推荐文章于 2023-10-15 12:21:54 发布

阅读量908

点赞数

CC 4.0 BY-SA版权

分类专栏： Grid 文章标签： os deployment server performance each

本文链接：https://blog.youkuaiyun.com/zhxue123/article/details/5277008

Grid 专栏收录该内容

21 篇文章

订阅专栏

本文介绍了两种用于高性能服务器集群的操作系统部署方式：一是通过部署软件如System Image等直接安装；二是利用NFS服务器进行集中存储并共享给各节点，后者又细分为单独目录配置和为每个节点复制系统实例的方法。

1) deploy OS in every node

You can use some deployment software to do that, such as SystemImage, etc. One of my paper 'study of transfer model of high performance server cluster'details something of this method.

2) NFS

2.1) You can copy the OS into a NFS server, and share it to all servers, but you need seperate some files/directory from each other. i.e., you need create some dir for each node. such as /tmp /var /etc

2.2) You can copy a OS into a NFS server, and duplicate it for each node. (supercomputing center adopts this method).