Hi Ralph
Thanks - i downloaded and installed openmpi-1.4a1r20435 and
now everything works as it should:
--output-filename : all processes write their outputs to the correct files
--xterm : all specified processes opened their xterms
I started my application with --xterm as i wrote in
Hi Jody
Well, the problem with both the output filename and the xterm option
was that I wasn't passing them back to the remote daemons under the
ssh launch environment. I should have that corrected now - things will
hopefully work with any tarball of r20407 or above.
Let me know...
Ralph
Ah! I know the problem - forgot you are running under ssh, so the
environment doesn't get passed.
I'll have to find a way to pass the output filename to the backend
nodes...should have it later today.
On Feb 3, 2009, at 11:09 AM, jody wrote:
Hi Ralph
--output-filename
It creates files,
Hi Ralph
>>
>> --output-filename
>> It creates files, but only for the local processes:
>> [jody@localhost neander]$ mpirun -np 8 -hostfile testhosts
>> --output-filename gnana ./MPITest
>> ... output ...
>> [jody@localhost neander]$ ls -l gna*
>> -rw-r--r-- 1 jody morpho 549 2009-02-03 18:02 gna
On Feb 3, 2009, at 10:16 AM, jody wrote:
Hi Ralph
Thanks for the fixes and the "!".
--xterm:
The "!" works, but i still don't have any xterms from my remote nodes
even with all my xhost+ and -x DISPLAY tricks explained below :(
--output-filename
It creates files, but only for the local proces
Hi Ralph
Thanks for the fixes and the "!".
--xterm:
The "!" works, but i still don't have any xterms from my remote nodes
even with all my xhost+ and -x DISPLAY tricks explained below :(
--output-filename
It creates files, but only for the local processes:
[jody@localhost neander]$ mpirun -np 8
On Feb 2, 2009, at 2:55 AM, jody wrote:
Hi Ralph
The new options are great stuff!
Following your suggestion, i downloaded and installed
http://www.open-mpi.org/nightly/trunk/openmpi-1.4a1r20392.tar.gz
and tested the new options. (i have a simple cluster of
8 machines over tcp). Not everything
Okay, I have this fixed and the man page updated as of r20396.
Thanks again for finding and reporting this bug!
Ralph
On Feb 2, 2009, at 5:55 AM, Ralph Castain wrote:
Hmnmm...well, it shouldn't crash (so I'll have to fix that), but it
should fail. The --report-pid option takes an argument,
Great feedback - thanks!
I'll look into these and see what's going on. We had tested things on
a couple of systems, but it sounds like maybe there are some system-
specific issues we are encountering. Let me explore a little to see
what might be happening.
Ralph
On Feb 2, 2009, at 2:55 AM
Hmnmm...well, it shouldn't crash (so I'll have to fix that), but it
should fail. The --report-pid option takes an argument, which wasn't
provided here. I'll check the man page to ensure it is up-to-date.
What it should tell you is that --report-pid takes either a '-' to
indicate that the pi
Hi Ralph
one more thing i noticed while trying out orte_iof again.
The option --report-pid crashes mpirun:
[jody@localhost neander]$ mpirun -report-pid -np 2 ./MPITest
[localhost:31146] *** Process received signal ***
[localhost:31146] Signal: Segmentation fault (11)
[localhost:31146] Signal code:
Hi Ralph
The new options are great stuff!
Following your suggestion, i downloaded and installed
http://www.open-mpi.org/nightly/trunk/openmpi-1.4a1r20392.tar.gz
and tested the new options. (i have a simple cluster of
8 machines over tcp). Not everything worked as specified, though:
* timestamp-ou
I'm afraid we discovered a bug in optimized builds with r20392. Please
use any tarball with r20394 or above.
Sorry for the confusion
Ralph
On Feb 1, 2009, at 5:27 AM, Jeff Squyres wrote:
On Jan 31, 2009, at 11:39 AM, Ralph Castain wrote:
For anyone following this thread:
I have completed
On Jan 31, 2009, at 11:39 AM, Ralph Castain wrote:
For anyone following this thread:
I have completed the IOF options discussed below. Specifically, I
have added the following:
* a new "timestamp-output" option that timestamp's each line of output
* a new "output-filename" option that redi
For anyone following this thread:
I have completed the IOF options discussed below. Specifically, I have
added the following:
* a new "timestamp-output" option that timestamp's each line of output
* a new "output-filename" option that redirects each proc's output to
a separate rank-named f
That's cool then - i have written a shellscript
which automatically does the xhost stuff for all
nodes in my hostfile :)
On Mon, Jan 26, 2009 at 9:25 PM, Ralph Castain wrote:
>
> On Jan 26, 2009, at 1:20 PM, jody wrote:
>
>> Hi Brian
>>
>>>
>>> I would rather not have mpirun doing an xhost comman
On Jan 26, 2009, at 1:20 PM, jody wrote:
Hi Brian
I would rather not have mpirun doing an xhost command - I think
that is
beyond our comfort zone. Frankly, if someone wants to do this, it
is up to
them to have things properly setup on their machine - as a rule, we
don't
mess with your
Typo there: "xceren" stands for "screen" - sorry :)
On Mon, Jan 26, 2009 at 9:20 PM, jody wrote:
> Hi Brian
>
>>
>> I would rather not have mpirun doing an xhost command - I think that is
>> beyond our comfort zone. Frankly, if someone wants to do this, it is up to
>> them to have things properly
Hi Brian
>
> I would rather not have mpirun doing an xhost command - I think that is
> beyond our comfort zone. Frankly, if someone wants to do this, it is up to
> them to have things properly setup on their machine - as a rule, we don't
> mess with your machine's configuration. Makes sys admins u
Hi Jody
I would rather not have mpirun doing an xhost command - I think that
is beyond our comfort zone. Frankly, if someone wants to do this, it
is up to them to have things properly setup on their machine - as a
rule, we don't mess with your machine's configuration. Makes sys
admins ups
Yes. The --tag-output option will prepend [job,rank] (or
stderr, whichever is appropriate) to each line. I don't insert a
colon, though I suppose that would easily be done for grep purposes.
I just finished implementing the --output-filename option that will
split the output from each rank
Hello Ralph:
Please forgive if this has already been covered...
Have you considered prefixing each line of output from each process
with something like "process_number" and a colon?
That is what IBM's poe does. Separating the output is then easy:
cat file | grep 0: > file.0
cat file | grep
On Sun, 2009-01-25 at 05:20 -0700, Ralph Castain wrote:
> 2. redirect output of specified processes to files using the provided
> filename appended with ".rank". You can do this for all ranks, or a
> specified subset of them.
A filename extension including both the comm size and the rank is
h
Hi
I have written some shell scripts which ease the output
to an xterm for each processor for normal execution(run_sh.sh),
gdb (run_gdb.sh), and valgrind (run_vg.sh).
In order for the xterms to be shown on your machine,
you have to set the DISPLAY variable on every host
(if this is not done by ssh
For those of you following this thread:
I have been impressed by the various methods used to grab the output
from processes. Since this is clearly something of interest to a broad
audience, I would like to try and make this easier to do by adding
some options to mpirun. Coming in 1.3.1 will
jody wrote:
Hi
I have a small cluster consisting of 9 computers (8x2 CPUs, 1x4 CPUs).
I would like to be able to observe the output of the processes
separately during an mpirun.
What i currently do is to apply the mpirun to a shell script which
opens a xterm for each process,
which then starts t
On Thu, 2009-01-22 at 06:33 -0700, Ralph Castain wrote:
> If you need to do this with a prior releasewell, I'm afraid it
> won't work. :-)
As a quick hack for 1.2.x, I sometimes use this script to wrap my
executable:
---
#!/bin/sh
#
Hmmmit works fine for me. Note that orte-iof must be executed from
the same node as mpirun. Also, ensure that rank=0 is actually the MPI
rank that is printing something! Otherwise, nothing will come out.
It should also be noted that it does take time for you to execute orte-
iof, and tha
Hi
I downloaded 1.3 the day its release was announced :)
I now wanted to try it, so i pieced it together from orte-iof --help:
- started an mpirun with 4 process
- opened a new terminal
- did ps to find the mpirun's pid
- then: orte-iof --pid 8449 --rank 0 --stdout
but nothing happens...
So did i
If you are willing/able to upgrade to 1.3, you will find a new utility
called "ompi-iof" (or "orte-iof") that allows you to capture the
output from any specified rank or combination of ranks. Note that the
output is copied to the tool, so it will also still arrive at the
output of mpirun.
30 matches
Mail list logo
