Interesting. It seems that the mpirun that you have executed is not
in the 7.0 series. Note that it doesn't have a -ssi parameter in the
help output. It should have a line like
-sigs / -nsigs Catch/don't catch signals in MPI
application
-ssi <n> <arg> Set environment variable
LAM_MPI_SSI_<n>=<arg>
-toff Enable tracing with generation
initially off
What is likely happening is that your path is not set (or being set)
correctly to point to the install of 7.1.2 that you did. the output
of the following three commands, in sequence should highlight this.
$ which lamboot
$ which mpirun
$ lamboot -V
$ which mpirun
Let me know if that helps.
Josh
On Apr 14, 2006, at 2:19 AM, Zhenxia Zhang wrote:
> Thanks.
> I do no have an alias setup for mpirun. Using the full path has the
> same result.
>
> The output of $ mpirun -h
> ----------------------------------------------------------------------
> ----------------
> Synopsis: mpirun [options] <app>
> mpirun [options] <where> <program> [<prog args>]
>
> Description: Start an MPI application in LAM/MPI.
>
> Notes:
> [options] Zero or more of the options listed
> below.
> <app> LAM/MPI appschema.
> <where> List of LAM nodes and/or CPUs
> (examples
> below).
> <program> Must be a LAM/MPI program that either
> invokes MPI_INIT or has exactly one of
> its children invoke MPI_INIT.
> <prog args> Optional list of command line
> arguments
> to <program>.
>
> Options:
> -c <num> Run <num> copies of <program> (same
> as -np).
> -c2c Use fast library (C2C) mode.
> -client <rank> <host>:<port>
> Run IMPI job; connect to the IMPI
> server <host>
> at port <port> as IMPI client
> number <rank>.
> -D Change current working directory of
> new
> processes to the directory where the
> executable resides.
> -f Do not open stdio descriptors.
> -ger Turn on GER mode.
> -h Print this help message.
> -lamd Use LAM daemon (LAMD) mode
> (opposite of -c2c).
> -nger Turn off GER mode.
> -np <num> Run <num> copies of <program> (same
> as -c)
> -nsigs Don't catch signals.
> -nx Don't export LAM_MPI_* environment
> variables.
> -O Multicomputer is homogeneous.
> -pty Use pty if stdout is a tty.
> -s <nodeid> Load <program> from node <nodeid>.
> -toff Enable tracing with generation
> initially off.
> -ton, -t Enable tracing with generation
> initially on.
> -v Be verbose.
> -w / -nw Wait/don't wait for application to
> complete.
> -wd <dir> Change current working directory of
> new
> processes to <dir>.
> -x <envlist> Export environment vars in <envlist>.
>
> Nodes: n<list>, e.g., n0-3,5
> CPUS: c<list>, e.g., c0-3,5
> Extras: h (local node), o (origin node), N (all nodes), C
> (all CPUs)
>
> Examples: mpirun n0-7 prog1
> Executes "prog1" on nodes 0 through 7.
>
> mpirun -lamd -x FOO=bar,DISPLAY N prog2
> Executes "prog2" on all nodes using the LAMD RPI.
> In the environment of each process, set FOO to the
> value
> "bar", and set DISPLAY to the current value.
>
> mpirun n0 N prog3
> Run "prog3" on node 0, *and* all nodes. This
> executes *2*
> copies on n0.
>
> mpirun C prog4 arg1 arg2
> Run "prog4" on each available CPU with command line
> arguments of "arg1" and "arg2". If each node has a
> CPU count of 1, the "C" is equivalent to "N". If at
> least one node has a CPU count greater than 1, LAM
> will run neighboring ranks of MPI_COMM_WORLD on that
> node. For example, if node 0 has a CPU count of 4 and
> node 1 has a CPU count of 2, "prog4" will have
> MPI_COMM_WORLD ranks 0 through 3 on n0, and ranks 4
> and 5 on n1.
>
> mpirun c0 C prog5
> Simlar to the "prog3" example above, this runs "prog5"
> on CPU 0 *and* on each available CPU. This executes
> *2* copies on the node where CPU 0 is (i.e., n0).
> This is probably not a useful use of the "C" notation;
> it is only shown here for an example.
>
> Defaults: -c2c -w -nger -pty
> ----------------------------------------------------------------------
> ----------------
>
> Is it related to the kernel version? I try the same commands on
> Fedro with kernel 2.6.11, that's OK.
>
> Sincerely,
> Zhenxia Zhang
>
> ----- Original Message -----
> From: "Josh Hursey" <jjhursey_at_[hidden]>
> To: "General LAM/MPI mailing list" <lam_at_[hidden]>
> Sent: Friday, April 14, 2006 3:02 AM
> Subject: Re: LAM: Problem about using checkpoint with blcr
>
>
>> Humm... That's interesting. Do you happen to have an alias setup for
>> mpirun? (the command 'alias mpirun' should return something if you
>> do.). Other than that I can only think of two things to try:
>> - Use the full path to the mpirun command when executing it:
>> $ /usr/local/bin/mpirun -ssi rpi crtcp C ./hello_mpi
>> - Send the output of:
>> $ mpirun -h
>> And let's see if that matches what is expected.
>>
>> Sorry I can't be much more help. :/
>>
>> Cheers,
>> Josh
>>
>> On Apr 13, 2006, at 2:44 AM, Zhenxia Zhang wrote:
>>
>>> I tried the command: $ mpirun -ssi rpi crtcp C ./hello_mpi again
>>> and check it carefully. But the result is same as before. Is there
>>> any other possibilities?
>>>
>>> Regards,
>>> Zhenxia Zhang
>>> _______________________________________________
>>> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>>
>> ----
>> Josh Hursey
>> jjhursey_at_[hidden]
>> http://www.lam-mpi.org/
>>
>> _______________________________________________
>> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>>
>
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
----
Josh Hursey
jjhursey_at_[hidden]
http://www.lam-mpi.org/
|
