Sponsored Content
Operating Systems Linux Rocks cluster 6.1 and MPICH2 problem?????? Post 302756249 by gabam on Tuesday 15th of January 2013 12:58:08 PM
Old 01-15-2013
Rocks cluster 6.1 and MPICH2 problem??????

Hey friends,
I am trying to execute a simple hello world in mpi on MPICH2 on Rocks cluster. here is the c source code.

Code:
 
#include <mpi.h>
#include <stdio.h>
int main( int argc, char ** argv )
{
 MPI_Init( NULL, NULL );
 int world_size;
 MPI_Comm_size( MPI_COMM_WORLD, &world_size );
 int world_rank;
 MPI_Comm_rank( MPI_COMM_WORLD, &world_rank );
 char processor_name[MPI_MAX_PROCESSOR_NAME];
 int name_len;
 MPI_Get_processor_name( processor_name, &name_len );
 printf( "Hello world from processor %s, rank %d" " out of %d processors\n", processor_name, world_rank, world_size );
 MPI_Finalize();
}


And I compile it like this.

Code:
/opt/mpich2/gnu/bin/mpicc ./hello.c -o hello

I have the following entry on the machine file.

Code:
compute-0-0
compute-0-1

Now here is how I run the hello program

Code:
/opt/mpich/gnu/bin/mpirun -np 2 -machinefile machines ./hello

which gives me the follwing error.

Code:
 
[user1@cluster ~]$ /opt/mpich2/gnu/bin/mpirun -np 2 -machinefile machines ./mpi_hello_world
Could not chdir to home directory /export/home/user1: No such file or directory
Could not chdir to home directory /export/home/user1: No such file or directory
[proxy:0:0@compute-0-0.local] launch_procs (./pm/pmiserv/pmip_cb.c:687): unable to change wdir to /export/home/user1 (No such file or directory)
[proxy:0:0@compute-0-0.local] HYD_pmcd_pmip_control_cmd_cb (./pm/pmiserv/pmip_cb.c:935): launch_procs returned error
[proxy:0:0@compute-0-0.local] HYDT_dmxu_poll_wait_for_event (./tools/demux/demux_poll.c:77): callback returned error status
[proxy:0:0@compute-0-0.local] [mpiexec@cluster.hpc.org] control_cb (./pm/pmiserv/pmiserv_cb.c:215): assert (!closed) failed
[mpiexec@cluster.hpc.org] HYDT_dmxu_poll_wait_for_event (./tools/demux/demux_poll.c:77): callback returned error status
[mpiexec@cluster.hpc.org] HYD_pmci_wait_for_completion (./pm/pmiserv/pmiserv_pmci.c:181): error waiting for event
[mpiexec@cluster.hpc.org] main (./ui/mpich/mpiexec.c:405): process manager error waiting for completion
[user1@cluster ~]$

Please help me.
 

4 More Discussions You Might Find Interesting

1. Linux

Intermittent connectivity issues with ROCKS on a compute cluster

I have a cluster set up with a head node and compute nodes running TORQUE and MOAB. The distro is ROCKS 5.3. I've been having problems with the connectivity for the past couple weeks now. Every couple hours it seems like the network connectivity will just stop working: sometimes it'll start back up... (0 Replies)
Discussion started by: gandalf85
0 Replies

2. UNIX for Dummies Questions & Answers

MPICH2 version/Error

Guys, I have two questions: 1- What command do i have to use to know the version of MPICH2? 2- I am running a parallel computation on quantum espresso using a SLURM interface (INTEL) and I am getting the following error: however a colleague gave me another (rather more complicated) script... (1 Reply)
Discussion started by: lebphys78
1 Replies

3. Red Hat

Cluster Problem

Hi, I am facing issuing in setting up Linux cluster. Here is the issue that i am facing. I have 2 Linux desktop and have following ip's and name: hitesh12-192.168.1.23 saanvi12-192.168.1.30 i enabled ricci service and have setup passwod as well.Enabled luci service as well. When... (0 Replies)
Discussion started by: hitesh1907
0 Replies

4. UNIX for Advanced & Expert Users

CentOS 6.8 with Rocks Cluster: ldconfig is not a symbolic link errors

Any help appreciated just logging in to this server which is a front end for Rocks Cluster 6.1.1. Getting the below errors: ldconfig ldconfig: /usr/lib/libX11.so.6 is not a symbolic link ldconfig: /usr/lib/libjpeg.so.62 is not a symbolic link ldconfig: /usr/lib/libpng12.so.0 is not a symbolic... (3 Replies)
Discussion started by: RobbieTheK
3 Replies
cmdeleteconf(1m)														  cmdeleteconf(1m)

NAME
cmdeleteconf - Delete either the cluster or the package configuration SYNOPSIS
cmdeleteconf [-f] [-v] [-c cluster_name] [[-p package_name]...] DESCRIPTION
cmdeleteconf deletes either the entire cluster configuration, including all its packages, or only the specified package configuration. If neither cluster_name nor package_name is specified, cmdeleteconf will delete the local cluster's configuration and all its packages. If the local node's cluster configuration is outdated, cmdeleteconf without any argument will only delete the local node's configuration. If only the package_name is specified, the configuration of package_name in the local cluster is deleted. If both cluster_name and pack- age_name are specified, the package must be configured in the cluster_name, and only the package package_name will be deleted. cmdelete- conf with only cluster_name specified will delete the entire cluster configuration on all the nodes in the cluster, regardless of the con- figuration version. The local cluster is the cluster that the node running the cmdeleteconf command belongs to. Only a superuser, whose effective user ID is zero (see id(1) and su(1)), can delete the configuration. To delete the cluster configuration, halt the cluster first. To delete a package configuration you must halt the package first, but you do not need to halt the cluster (it may remain up or be brought down). To delete the package VxVM-CVM-pkg (HP-UX only), you must first delete all packages with STORAGE_GROUP defined. While deleting the cluster, if any of the cluster nodes are powered down, the user can choose to continue deleting the configuration. In this case, the cluster configuration on the down node will remain in place and, therefore, be out of sync with the rest of the cluster. If the powered-down node ever comes up, the user should execute the cmdeleteconf command with no argument on that node to clean up the config- uration before doing any other Serviceguard command. Options cmdeleteconf supports the following options: -f Force the deletion of either the cluster configuration or the package configuration. -v Verbose output will be displayed. -c cluster_name Name of the cluster to delete. The cluster must be halted already, if intending to delete the cluster. -p package_name Name of an existing package to delete from the cluster. The package must be halted already. There should not be any packages in the cluster with STORAGE_GROUP defined before having a package_name of VxVM-CVM-pkg (HP-UX only). RETURN VALUE
Upon completion, cmdeleteconf returns one of the following values: 0 Successful completion. 1 Command failed. EXAMPLES
The high availability environment contains the cluster, clusterA , and a package, pkg1. To delete package pkg1 in clusterA, do the following: cmdeleteconf -f -c clusterA -p pkg1 To delete the cluster clusterA and all its packages, do the following: cmdeleteconf -f -c clusterA AUTHOR
cmdeleteconf was developed by HP. SEE ALSO
cmcheckconf(1m), cmapplyconf(1m), cmgetconf(1m), cmmakepkg(1m), cmquerycl(1m). Requires Optional Serviceguard Software cmdeleteconf(1m)
All times are GMT -4. The time now is 01:30 AM.
Unix & Linux Forums Content Copyright 1993-2022. All Rights Reserved.
Privacy Policy