I'm looking for some help with mpich (parallel processing).  I've 
got task-parallel-computing-node installed on two connected machines 
(ivan and alexi), and task-parallel-computing-dev installed on 
one of them (ivan).  rsh-server is set up on both, and I've got 
a ~/.rhosts file on each that allows me into both machines.  I have 
also set up the /etc/mpich/machines.LINUX with the fully qualified 
domain names for the two machines.  Finally, my home directory is 
cross-mounted on both machines.  The whole thing is on a private
network behind a firewall machine.

Here's what I've tried (from the front-end machine):

    [EMAIL PROTECTED]:~$ tstmachines -v
    Trying true on ivan.tuxnet.edu ...
    Trying true on alexi.tuxnet.edu ...
    Trying ls on ivan.tuxnet.edu ...
    Trying ls on alexi.tuxnet.edu ...
    Trying user program on ivan.tuxnet.edu ...
    Trying user program on alexi.tuxnet.edu ...

    [EMAIL PROTECTED]:~$ cp /usr/share/doc/mpich/examples/code/Makefile.in ~/
    [EMAIL PROTECTED]:~$ cp /usr/share/doc/mpich/examples/code/cpi.c ~/

    [EMAIL PROTECTED]:~$ mpireconfig Makefile
    creating Makefile

    [EMAIL PROTECTED]:~$ make cpi
    /usr/lib/mpich/build/LINUX/ch_p4/bin/mpicc -O2 -c cpi.c
    /usr/lib/mpich/build/LINUX/ch_p4/bin/mpicc -O2 -o cpi cpi.o -lm

    [EMAIL PROTECTED]:~$ mpirun -np 2 cpi

At this point top shows on my front-end (ivan):

  PID USER     PRI  NI  SIZE  RSS SHARE STAT  LIB %CPU %MEM   TIME COMMAND
  634 cswingle  16   0   676  668   580 R       0 99.6  0.5   1:18 mpirun

It never seems to run cpi, and there are no rshd connections on my 
other machine (alexi), nor does mpirun or cpi every show up on alexi.  
mpirun just seems to hang.  Looking at the source code leads me to 
suspect that the program should immediately be asking me for the 
number of intervals to use in the calculation, but I never get that far.

I've tried several of the other programs in the examples/code directory,
but they either don't compile correctly, or exhibit the same freeze when
I do mpirun.

Does anyone have any hints on what I'm doing wrong, or what I should
try next?

Thanks.

Chris
-- 
Christopher S. Swingley           tel: 907-474-2689 cell: 322-1889
930 Koyukuk Drive, Suite 408C     email: [EMAIL PROTECTED]
University of Alaska Fairbanks    www.frontier.iarc.uaf.edu:8080/
Fairbanks, AK  99775                     ~cswingle

PGP2 key:   http://www.frontier.iarc.uaf.edu:8080/~cswingle/pgpkey.asc
GNUPG key: http://www.frontier.iarc.uaf.edu:8080/~cswingle/gnupgkey.asc

Reply via email to