[mvapich-discuss] mpirun_rsh: Unable to get host entry

[Jonathan Perkins]

Hi, it looks like there is some problem with resolution of the
hostname.  Perhaps the compute node is unable to determine how to
reach the launching node.  This may be due to one of the components
believing that it can reach the other by using the loopback device.

Is your launching node the same as the node its running on.  If not,
can each node resolve the hostname of each other (either dns or
/etc/hosts)?  There may be an issue where the hostname of the machine
is assigned to the loopback interface on one of the machines in
/etc/hosts.  If this is the case you can edit the /etc/hosts file so
that the hostname is placed on the line for the external ip address of
the machine instead.

On Thu, Jul 21, 2011 at 10:59 PM, Champagne, Nathan J. (JSC-EV)[Jacobs
Technology] <nathan.j.champagne at nasa.gov> wrote:
> Hello,
> I'm using MVAPICH 1.7rc1 on a Gentoo Linux cluster with a Mellanox
> Infiniband switch. To test the installation, I am trying to run the cxxpi
> example as follows:
>
> $ mpirun_rsh -np 1 -hostfile ./hosts_test ./cxxpi
> [unset]: Unable to get host entry for
> [unset]: Unable to connect to  on 46570
> Fatal error in MPI_Init:
> Other MPI error
> [juggernaut:mpispawn_0][child_handler] MPI process (rank: 0, pid: 22152)
> exited with status 1
> $
>
> The file "hosts_test" contains the host IP addresses.
> Does anyone have any ideas that come to mind for what's happening?
> Thanks,
> Nathan
>
> _______________________________________________
> mvapich-discuss mailing list
> mvapich-discuss at cse.ohio-state.edu
> http://mail.cse.ohio-state.edu/mailman/listinfo/mvapich-discuss
>
>



-- 
Jonathan Perkins
http://www.cse.ohio-state.edu/~perkinjo