标签：running path 运行 class 成功 master 共享文件夹 htm 计算
在使用MPI框架中，需要多机进行通信进行并行计算；现在配置多个主机进行运行mpi程序，并进行通信；涉及到ssh无密码和nfs文件系统配置；
参考文档：
多机配置步骤：
运行状态：可见已经在两个节点上运行；
mpirun -n 10 -hosts client,master ./mpi_hello_world Hello world from processor jn-1, rank 8 out of 10 processors Hello world from processor jn-1, rank 6 out of 10 processors Hello world from processor jn-1, rank 4 out of 10 processors Hello world from processor jn-1, rank 0 out of 10 processors Hello world from processor jn-1, rank 2 out of 10 processors Hello world from processor jn-4, rank 5 out of 10 processors Hello world from processor jn-4, rank 7 out of 10 processors Hello world from processor jn-4, rank 1 out of 10 processors Hello world from processor jn-4, rank 3 out of 10 processors Hello world from processor jn-4, rank 9 out of 10 processors
节点host样本：主机要测试主从节点网络互相连通，此配置基于以太网配置；
$ cat /etc/hosts (master) 127.0.0.1 localhost #127.0.1.1 1944 #MPI CLUSTER SETUP 172.50.88.22 master 172.50.88.56 slave1 172.50.88.34 slave2
$ cat /etc/hosts (client) 127.0.0.1 localhost #127.0.1.1 1947 (在超算中，也经常使用数据进行编号节点) #MPI CLUSTER SETUP 172.50.88.22 master 172.50.88.54 slave3
保持更新，转载请注明出处；
标签：running path 运行 class 成功 master 共享文件夹 htm 计算
原文地址：https://www.cnblogs.com/xuyaowen/p/mpi-lan-run.html
