Dear OpenMPI prophets, I am running the OpenFOAM using the OpenMPI-2.1.1 on CentOS 7. The OpenMPI has been compiled by using the GCC-4.8.5. The command is something like this:
$ mpirun -n 256 -hostfile machines interFoam -parallel And the machines file contains the correct node information as follows: machines: ----------------- hpcnode001 16 hpcnode002 16 hpcnode003 16 ... I have sufferred from the low computation speed when I run the application by using mpirun command. When I ran the top command, I saw that the SYSTEM CPU USAGE takes too high values. i.e., Tasks: 418 total, 21 running, 398 sleeping, 0 stopped, 0 zombie %Cpu0 : 59.7 us, 40.3 sy, 0.0 ni, 0.0 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st %Cpu1 : 72.4 us, 27.6 sy, 0.0 ni, 0.0 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st %Cpu2 : 68.2 us, 31.8 sy, 0.0 ni, 0.0 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st ... Why is it happened? To tackle this problem, where should I start to check or monitor the system? Looking forward to hearing your opinion. Thank you. Regards, GEON-HONG KIM Engineer Hyundai Heavy Industries Co., Ltd. M. +82 10 3084 1357 E. kgb3...@gmail.com / geonh...@hhi.co.kr
