failed to handshake with mpd on hpc004; recvd output={}

failed to handshake with mpd on hpc004; recvd output={}

Dear Sir,

I cannot launch the mpdboot in the master
I have 2 nodes other than master node.
when I try to run mpdboot -r ssh -f mpd.hosts -n 3 -d I have been getting below error.
Could you please help me
Thanks in Advance.

debug: starting
running mpdallexit on hpcmas01.kfupm.edu.sa
debug: launch cmd= /home/atsheikh/mpich2-install/bin/mpd.py --ncpus=1 -e -d
debug: mpd on hpcmas01.kfupm.edu.sa on port 36586
debug: info for running mpd: {'ncpus': 1, 'list_port': 36586, 'entry_port': '', 'host': 'hpcmas01.kfupm.edu.sa', 'entry_host': '', 'ifhn': ''}
debug: launch cmd= ssh -x -n -q hpc004 '/home/atsheikh/mpich2-install/bin/mpd.py -h hpcmas01.kfupm.edu.sa -p 36586 --ncpus=1 -e -d'
debug: launch cmd= ssh -x -n -q hpc005 '/home/atsheikh/mpich2-install/bin/mpd.py -h hpcmas01.kfupm.edu.sa -p 36586 --ncpus=1 -e -d'
debug: mpd on hpc004 on port 49922
mpdboot_hpcmas01.kfupm.edu.sa (handle_mpd_output 406): failed to handshake with mpd on hpc004; recvd output={}

[1]+ Done mpd
Regards
Ashraf

2 posts / 0 new
Last post
For more complete information about compiler optimizations, see our Optimization Notice.

Hi Ashraf,

Are you using MPICH-2?
If you are going to use Intel MPI Library you need to set environment variable variables first.
Run:
'. /full/path/to/intel/mpi/installation/bin64/mpivars.sh'

Kill all existing mpd.py processes left from previous run.
And try one more time to run mpirun.

PS: you need to have passwordless access to all nodes.

Regards!
Dmitry

Leave a Comment

Please sign in to add a comment. Not a member? Join today