Hello everyone, I downloaded openmpi-1.3 version from night
tarballs to check RDMA-CM support. I am able to compile and install it, but
dont know how to run it as there is no documentation provided. Did someone try
running it with OpenMPI?My another question is Does OpenMPI1.3 has
progress-t
On Jun 17, 2008, at 5:12 AM, Neeraj Chourasia wrote:
I downloaded openmpi-1.3 version from night tarballs to check
RDMA-CM support. I am able to compile and install it, but dont know
how to run it as there is no documentation provided. Did someone try
running it with OpenMPI?
The them
This is pretty odd.
What files do you see in $prefix/share/openmpi/mpi*txt?
What do you see from:
grep required_file $prefix/share/openmpi/mpi*txt
On Jun 16, 2008, at 2:12 PM, Weirs, V Gregory wrote:
Dave Gunter:
According to ./configure —help, by default —enable-mpi-f77 and —
enabl
Is /tmp writable on your compute nodes?
On Jun 16, 2008, at 1:49 PM, Tony Smith wrote:
Dear Sir:
thanks.
I have changed it to its absolute path:
/ptmp/myname/openmpi123/ompi123_install/bin/mpirun -np 8 /ptmp/
myname/openmpi123/openmpi-1.2.3/examples/hello_c
But I still got the error :
==
On Mon, Jun 16, 2008 at 9:46 PM, Brock Palen wrote:
> Brad just curious.
> Did you tweak any other values for starting and running a job on such a
> large system? You say unmodified, but OpenMPI lets you tweak many values
> at runtime.
>
Ahh...sorry for the confusion. By "unmodified", I meant
On Jun 17, 2008, at 10:18 AM, Brad Benton wrote:
The main run-time parameters used are as follows:
- For scalable process launch
routed = binomial
Note that this is a new parameter for OMPI v1.3 -- it has to do with
how non-MPI communications are sent around (e.g., OMPI startup
coordina
thanks,
I changed /tmp and /ptmp and its sub directories to writable .
But , the same errors.
thanks,
> From: jsquy...@cisco.com
> To: us...@open-mpi.org
> Date: Tue, 17 Jun 2008 09:10:18 -0400
> Subject: Re: [OMPI users] help me please, about Open MPI
>
> Is /tmp writable on your compute nod
What are the exact permissions on /tmp? They should likely be 1777.
Do you have the TMPDIR or TMP environment variables set? If so, is
that directory also world-writable? (if set, these will override the
default location of /tmp)
The error you are seeing (session directory failed) usuall
I'm afraid that we've never tested OMPI under these conditions; I
don't know exactly what xgcc is testing for, so I don't know why it
would fail. There could be weird interactions between the OFED stack
and xgcc...? (i.e., memory "appears" from the kernel via ibv_*
function calls, etc.)
Hi Jeff,
i compiled OpenMPi using xgcc 4.0.2. I noted that the program stops before
to do process spawn.
Very strange.
2008/6/17 Jeff Squyres :
> I'm afraid that we've never tested OMPI under these conditions; I don't
> know exactly what xgcc is testing for, so I don't know why it would fail.
>
Hi
Since i upgraded from open-mpi .2.2 to open-mpi 1.2.5 and
had to reinstall my machine aim-plankton (fedora 8 instead of fedora 6)
open-mpi doesn't work correctly anymore:
When i start an application from aim-plankton to run on an other machine,
it seems to hang (no output, not even from debug-d
Hi Pat
A friendly elf forwarded this to me, so please be sure to explicitly include
me on any reply.
Was that the only error message you received? I would have expected a trail
of "error_log" outputs that would help me understand where this came from.
If not, I can give you some debug flags to se
Hi
OpenMPI does not pick up the infiniband component on our nodes with
Mellanox cards:
ompi_info --param btl openib
returns
MCA btl: parameter "btl_base_debug" (current value: "0")
If btl_base_debug is 1 standard debug is output, if > 1 verbose
debug is output
MCA btl: parameter "btl"
As you surmised, it looks like your "btl" parameter is being set
somewhere, and it excludes the openib BTL.
Check out this FAQ entry:
http://www.open-mpi.org/faq/?category=tuning#setting-mca-params
On Jun 17, 2008, at 10:49 AM, SLIM H.A. wrote:
Hi
OpenMPI does not pick up the infinib
thanks,
drwxrwxrwt9 root root102400 Jun 17 10:36 tmp
I have no TMPDIR or TMP environment variables set .
SO, do I need to set up some for them ?
thanks a lot !
June 17 2008
> From: jsquy...@cisco.com
> To: us...@open-mpi.org
> Date: Tue, 17 Jun 2008 10:35:06 -0400
> Subject: Re: [
thanks,
My configure :
./configure --prefix=/ptmp/myname/openmpi --enable-static --disable-shared
CC=icc CXX=icpc F77=ifort FC=ifort --with-gm=/opt/gm
-with-tm=/usr/spool/PBS/is that correct ?
thanks
> From: jsquy...@cisco.com
> To: us...@open-mpi.org
> Date: Tue, 17 Jun 2008 10:35:06 -040
Sure, that configure line should be fine. But that's a different
issue than the permissions on your temp directories. The environment
variables I was asking about are what are set at run time -- not when
you configure/build OMPI.
If you're using Torque, I *believe* that it sets TMPDIR in
I have
mpic++-wrapper-data.txt
mpicc-wrapper-data.txt
mpicxx-wrapper-data.txt
mpif77-wrapper-data.txt
mpif90-wrapper-data.txt
And for the grep:
mpic++-wrapper-data.txt:required_file=
mpicc-wrapper-data.txt:required_file=
mpicxx-wrapper-data.txt:required_file=
mpif77-wrapper-data.txt:required_fi
Jeff,
Thanks, I forgot I had a ~/.openmpi/mca-params.conf file laying around
from a previous test.
Henk
> -Original Message-
> From: users-boun...@open-mpi.org
> [mailto:users-boun...@open-mpi.org] On Behalf Of Jeff Squyres
> Sent: 17 June 2008 16:03
> To: Open MPI Users
> Subject: Re:
On Jun 6, 2008, at 6:03 AM, Gabriele Fatigati wrote:
Hi Jeff,
Sorry for the delay in replying -- I was on vacation all last week.
thanks for you reply. I did understand previous questions about
RDMA. Ever with SKaMPI, i tried to run with mpi_leave_pinned = 1, as
you have suggested. But a
Thanks for digging into this!
The assembly portion of OMPI is quite squirrelly and dangerous to mess
with. We'll need to check into this carefully to make sure that it
works properly on all supported architectures...
As for other bounds checking, would you mind checking the OMPI
developm
I can believe 1.2.x has problems in that regard. Some of that has nothing to
do with slurm and reflects internal issues with 1.2.
We have made it much more resistant to those problems in the upcoming 1.3
release, but there is no plan to retrofit those changes to 1.2. Part of the
problem was that w
thanks,
I added "setenv TMPDIR /tmp:$TMPDIR" in my job script file.
so , echo $TMPDIR :
/tmp:/local2/pbs/myname/37911.hpc-cluster
but the same errors.
thanks
> From: jsquy...@cisco.com
> To: us...@open-mpi.org
> Date: Tue, 17 Jun 2008 11:55:05 -0400
> Subject: Re: [OMPI users] help me
TMPDIR should be a single directory:
export TMPDIR=/tmp
It should not be a list of directories seperated by a ':'.
"setenv TMPDIR /tmp" (or "setenv TMPDIR
/local2/pbs/myname/37911.hpc-cluster") would hence be the correct line.
After setting the variable, try
'cd $TMPDIR'
'touch test'
to veri
On Tue, Jun 17, 2008 at 01:11:15PM -0400, Tony Smith wrote:
> so , echo $TMPDIR :
>
> /tmp:/local2/pbs/myname/37911.hpc-cluster
>
> but the same errors.
Have you checked "ls -ld /tmp/"? Jeff has been asking you twice to
provide the exact permissions, so I guess he would be delighted to
ev
On Tue, Jun 17, 2008 at 07:30:27PM +0200, Adrian Knoth wrote:
> Have you checked "ls -ld /tmp/"? Jeff has been asking you twice to
Never mind, I just saw your other mail (tagged with [SPAM]) ;)
--
Cluster and Metacomputing Working Group
Friedrich-Schiller-Universität Jena, Germany
private: ht
Alright, I'm an idiot. I didn't adjust my DYLD_LIBRARY_PATH.
Sorry about that-
Greg
On 6/17/08 7:08 AM, "Jeff Squyres" wrote:
This is pretty odd.
What files do you see in $prefix/share/openmpi/mpi*txt?
What do you see from:
grep required_file $prefix/share/openmpi/mpi*txt
On Jun 16, 2
Sorry for not replying earlier -- glad you figured it out.
Yes, if you didn't also reset the LIBRARY_PATH, the "wrong" libmpi
(and friends) will get picked up, and it will therefore use the hard-
coded directory path in those "wrong" libraries to find the wrapper
data text files (the ones wi
28 matches
Mail list logo
