This article shows how to run the Intel MPI Benchmark (IMB) on a Bright cluster. In this example we will run the benchmark over low latency 10GbE interfaces.
Here's a simple example of how to run an OpenMPI program using openmpi in Bright Cluster Manager, through the Slurm workload manager.
Write a job script.
$ cat hello.sh
#!/bin/sh
#SBATCH -o hello.out
#SBATCH --nodes=3
#SBATCH --ntasks-per-node=2
mpirun hello
$ sbatch hello.sh
Submitted batch job 35
$ squeue
JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON)
35 defq hello.sh rstober R 0:04 3 atom[01-03]
$ scontrol show job 35
JobId=35 Name=hello.sh
UserId=rstober(1001) GroupId=rstober(1001)
Priority=4294901758 Account=(null) QOS=normal
JobState=COMPLETED Reason=None Dependency=(null)
Requeue=1 Restarts=0 BatchFlag=1 ExitCode=0:0
RunTime=00:00:02 TimeLimit=UNLIMITED TimeMin=N/A
SubmitTime=2012-07-17T07:41:39 EligibleTime=2012-07-17T07:41:39
StartTime=2012-07-17T07:41:39 EndTime=2012-07-17T07:41:41
PreemptTime=None SuspendTime=None SecsPreSuspend=0
Partition=defq AllocNode:Sid=atom-head1:24006
ReqNodeList=(null) ExcNodeList=(null)
NodeList=atom[01-03]
BatchHost=atom01
NumNodes=3 NumCPUs=10 CPUs/Task=1 ReqS:C:T=*:*:*
MinCPUsNode=2 MinMemoryNode=0 MinTmpDiskNode=0
Features=(null) Gres=(null) Reservation=(null)
Shared=0 Contiguous=0 Licenses=(null) Network=(null)
Command=/home/rstober/slurmhello.sh
WorkDir=/home/rstober
$ squeue
JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON)
$ more hello.out
Hello world from process 001 out of 006, processor name atom01
Hello world from process 002 out of 006, processor name atom02
Hello world from process 003 out of 006, processor name atom02
Hello world from process 000 out of 006, processor name atom01
Hello world from process 004 out of 006, processor name atom03
Hello world from process 005 out of 006, processor name atom03![]()