good idea anyway. It hurts nothing, takes milliseconds
to do, and guarantees nothing got left behind (e.g., if someone was
using a debug version of OMPI and directed opal_output to a file).
On Nov 4, 2011, at 4:43 AM, TERRY DONTJE wrote:
David, are you saying your jobs consistently leave behind
On 11/23/2011 2:02 PM, Paul Kapinos wrote:
Hello Ralph, hello all,
Two news, as usual a good and a bad one.
The good: we believe to find out *why* it hangs
The bad: it seem for me, this is a bug or at least undocumented
feature of Open MPI /1.5.x.
In detail:
As said, we see mystery hang-ups
Are all the other processes gone? What version of OMPI are you using?
On 11/28/2011 9:00 AM, Mudassar Majeed wrote:
Dear people,
In my MPI application, all the processes call
the MPI_Finalize (all processes reach there) but the rank 0 process
could not finish with MPI_F
IIRC, RNR's are usually due to the receiving side not having a segment
registered and ready to receive data on a QP. The btl does go through a
big dance and does its own flow control to make sure this doesn't happen.
So when this happens are both the sending and receiving nodes using
mthca's
w.
On Thu, Dec 15, 2011, at 07:00 AM, TERRY DONTJE wrote:
IIRC, RNR's are usually due to the receiving side not having a segment
registered and ready to receive data on a QP. The btl does go through a
big dance and does its own flow control to make sure this doesn't happen.
So when this
I am a little confused by your problem statement. Are you saying you
want to have each MPI process to have multiple threads that can call MPI
concurrently? If so you'll want to read up on the MPI_Init_thread
function.
--td
On 1/11/2012 7:19 AM, Hamilton Fischer wrote:
Hi, I'm actually usin
Do you have a stack of where exactly things are seg faulting in
blacs_pinfo?
--td
On 1/13/2012 8:12 AM, Conn ORourke wrote:
Dear Openmpi Users,
I am reserving several processors with SGE upon which I want to run a
number of openmpi jobs, all of which individually (and combined) use
less tha
Is there a way to set up an interface analogous to Unix's loopback? I
suspect setting "-mca btl self,sm" wouldn't help since this is probably
happening while the processes are bootstrapping.
--td
On 1/16/2012 7:26 PM, Ralph Castain wrote:
The problem is that OMPI is looking for a tcp port fo
ompi_info should tell you the current version of Open MPI your path is
pointing to.
Are you sure your path is pointing to the area that the OpenFOAM package
delivered Open MPI into?
--td
On 1/27/2012 5:02 AM, Brett Tully wrote:
Interesting. In the same set of updates, I installed OpenFOAM from
Do you get any interfaces shown when you run "ibstat" on any of the
nodes your job is spawned on?
--td
On 2/15/2012 1:27 AM, Tohiko Looka wrote:
Mm... This is really strange
I don't have that service and there is no ib* output in 'ifconfig -a'
or 'Infinband' in 'lspci'
Which makes me believe
To determine if an MPI process is waiting for a message do what Rayson
suggested and attach a debugger to the processes and see if any of them
are stuck in MPI. Either internally in a MPI_Recv or MPI_Wait call or
looping on a MPI_Test call.
Other things to consider.
Is this the first time y
and Infiniband,
*
--td
**
On Tue, Apr 24, 2012 at 3:02 PM, TERRY DONTJE <mailto:terry.don...@oracle.com>> wrote:
To determine if an MPI process is waiting for a message do what
Rayson suggested and attach a debugger to the processes and see if
any of them are stuck in MPI
On 4/25/2012 1:00 PM, Jeff Squyres wrote:
On Apr 25, 2012, at 12:51 PM, Ralph Castain wrote:
Sounds rather bizarre. Do you have lstopo on your machine? Might be useful to
see the output of that so we can understand what it thinks the topology is like
as this underpins the binding code.
The
On 5/4/2012 8:26 AM, Rolf vandeVaart wrote:
2. If that works, then you can also run with a debug switch to see
what connections are being made by MPI.
You can see the connections being made in the attached log:
[archimedes:29820] btl: tcp: attempting to connect() to [[60576,1],2] address
13
On 5/4/2012 1:17 PM, Don Armstrong wrote:
On Fri, 04 May 2012, Rolf vandeVaart wrote:
On Behalf Of Don Armstrong
On Thu, 03 May 2012, Rolf vandeVaart wrote:
2. If that works, then you can also run with a debug switch to
see what connections are being made by MPI.
You can see the connections
On 5/7/2012 8:40 PM, Jeff Squyres (jsquyres) wrote:
On May 7, 2012, at 8:31 PM, Jingcha Joba wrote:
So in the above stated example, end-start will be: +
20ms ?
(time slice of P2 + P3 = 20ms)
More or less (there's nonzero amount of time required for the kernel scheduler,
and the time quantu
BTW, the changes prior to r26496 failed some of the MTT test runs on
several systems. So if the current implementation is deemed not
"correct" I suspect we will need to figure out if there are changes to
the tests that need to be done.
See http://www.open-mpi.org/mtt/index.php?do_redir=2066 f
This looks like a missing check in the sctp configure.m4. I am working
on a patch.
--td
On 6/5/2012 10:10 AM, Siegmar Gross wrote:
Hello,
I compiled "openmpi-1.6" on "Solaris 10 sparc" and "Solaris 10 x86"
with "gcc-4.6.2" and "Sun C 5.12". Today I searched my log-files for
"WARNING" and fou
On 6/6/2012 4:38 AM, Siegmar Gross wrote:
Hello,
I compiled "openmpi-1.6" on "Solaris 10 sparc", "Solaris 10 x86",
and Linux (openSuSE 12.1) with "Sun C 5.12". Today I searched my
log-files for "WARNING" and found the following message.
WARNING: ***
Can you get on one of the nodes and see the job's processes? If so can
you then attach a debugger to it and get a stack? I wonder if the
processes are stuck in MPI_Init?
--td
On 6/7/2012 6:06 AM, Duke wrote:
Hi again,
Somehow the verbose flag (-v) did not work for me. I tried
--debug-daem
Another sanity think to try is see if you can run your test program on
just one of the nodes? If that works more than likely MPI is having
issues setting up connections between the nodes.
--td
On 6/7/2012 6:06 AM, Duke wrote:
Hi again,
Somehow the verbose flag (-v) did not work for me. I tr
his being a firewall issue is something
to look into.
--td
On 6/7/2012 6:36 AM, Duke wrote:
On 6/7/12 5:31 PM, TERRY DONTJE wrote:
Can you get on one of the nodes and see the job's processes? If so
can you then attach a debugger to it and get a stack? I wonder if
the processes are st
On 6/16/2012 8:03 AM, Roland Schulz wrote:
Hi,
I would like to start a single process without mpirun and then use
MPI_Comm_spawn to start up as many processes as required. I don't want
the parent process to take up any resources, so I tried to disconnect
the inter communicator and then finali
I am not sure I am understanding the problem correctly so let me
describe it back to you with a couple clarifications.
So your program using sf_open compiles successfully when using gcc and
mpicc. However, when you run the executable compiled using mpicc
sndFile is null?
If the above is rig
sing to the compiler.
This should give you an idea the difference between your gcc and mpicc
compilation. I would suspect either mpicc is using a compiler
significantly different than gcc or that mpicc might be passing some
optimization parameter that is messing the code execution (just a gu
I have
to make this program. Please have a look in a picture from link below,
maybe it will be more clear.
http://vipjg.nazwa.pl/sndfile_error.png
2012/7/30 TERRY DONTJE:
On 7/30/2012 6:11 AM, Paweł Jaromin wrote:
Hello
Thanks for fast answer, but the problem looks a little different.
Do you know what r# of 1.6 you were trying to compile? Is this via the
tarball or svn?
thanks,
--td
On 7/30/2012 9:41 AM, Daniel Junglas wrote:
Hi,
I compiled OpenMPI 1.6 on a 64bit Solaris ultrasparc machine.
Compilation and installation worked without a problem. However,
when trying to ru
Hi Hammad,
It looks to me like none of the btl's could resolve a route between the
node that process rank 0 is on to the other nodes.
I would suggest trying np=2 over a couple pairs of machines to see if
that works and you can truly be sure that only the
first node is having this problem.
It
Marcin,
A couple questions:
What OS are you running on?
Did you run this job oversubscribed, that is more processes than there
are cpus?
I've found with oversubscribed jobs that the recursive calls to
opal_progress by the SM BTL that
the yield within opal_progress (intending to give up the c
What kind of system lib errors are you seeing and do you have a stack
trace? Note, I was trying something similar with Solaris and 64-bit on
a SPARC machine and was seeing segv's inside the MPI Library due to a
pointer being passed through an integer (thus dropping the upper 32
bits). Funny t
e on SPARC except
the address is smaller on the former.
Greg, I would be interested in knowing if you are still seeing the
problem on Leopard and whether the above setting helps any.
--td
*
*Subject:* Re: [OMPI users] ScaLapack and BLACS on Leopard
*From:* Terry Dontje (/Terry.Dontje_at_[h
WHATMPI.
Sorry for the misinformation,
--td
Terry Dontje wrote:
Ok, I think I found the cause of the SPARC segv when trying to use a
64-bit compiled Open MPI library. If one does not set the WHATMPI
variable in the Bmake.inc it defaults to UseF77Mpi which assumes all
handles are ints. This is a
drop another
email with the results and I'll see what I can do.
hth,
Terry Dontje
Sun Microsystems, Inc.
Date: Mon, 21 Apr 2008 11:01:50 -0400
From: cfdm...@aim.com
Subject: [OMPI users] Problem with Sun Fortran Install with OpenMPI
To: us...@open-mpi.org
Message-ID: <8ca71d7c32c3
eciate the help,
Rob
-Original Message-----
From: Terry Dontje
To: us...@open-mpi.org
Sent: Mon, 21 Apr 2008 11:36 am
Subject: Re: [OMPI users] Problem with Sun Fortran Install with OpenMPI
Looking at the gcc.error attachment that looks to be the one of the
problems talked about
You can translate the communicator from Fortran to C using the MPI_COMM_F2C
routine.
--td
Message: 4
Date: Thu, 05 Jun 2008 08:53:55 +0200
From: Samuel Sarholz
Subject: [OMPI users] Communicators in Fortran and C
To: us...@open-mpi.org
Message-ID: <48478d83.6080...@rz.rwth-aachen.de>
Content-T
I have not tested this type of setup so the following disclaimer needs to be said. These are not exactly the same release number. They are close but their code could have something in them that makes them incompatible.
One idea comes to mind is whether the two nodes are on the same subnet? If
ly out of the
hands of the TCP stack and lower).
Alexander Shabarshin P.S. Between Linuxes I even tried different
versions of OpenMPI 1.2.4 and 1.2.5 - these versions work together
correctly, but not with ClusterTools...
Are the linuxes boxes on the same subnet?
--td
- Original Message - F
Date: Tue, 29 Jul 2008 14:19:14 -0400
From: "Alexander Shabarshin"
Subject: Re: [OMPI users] Communitcation between OpenMPI and
ClusterTools
To:
Message-ID: <00b701c8f1a7$9c24f7c0$c8afcea7@Shabarshin>
Content-Type: text/plain; format=flowed; charset="iso-8859-1";
reply-type=resp
Terry Dontje wrote:
Date: Tue, 29 Jul 2008 14:19:14 -0400
From: "Alexander Shabarshin"
Subject: Re: [OMPI users] Communitcation between OpenMPI and
ClusterTools
To:
Message-ID: <00b701c8f1a7$9c24f7c0$c8afcea7@Shabarshin>
Content-Type: text/plain; format=flowed; cha
really do expect this to work.
--td
Terry Dontje wrote:
Terry Dontje wrote:
Date: Tue, 29 Jul 2008 14:19:14 -0400
From: "Alexander Shabarshin"
Subject: Re: [OMPI users] Communitcation between OpenMPI and
ClusterTools
To:
Message-ID: <00b701c8f1a7$9c24f7c0$c8afcea7@Shabarshin&
Interestingly enough on the SPARC platform the Solaris memcpy's actually
use non-temporal stores for copies >= 64KB. By default some of the mca
parameters to the sm BTL stop at 32KB. I've done experimentations of
bumping the sm segment sizes to above 64K and seen incredible speedup on
our M90
); different compiler authors have chosen to
implement different optimizations that work well in different
applications. So yes, you may well see different run-time performance
with different compilers depending on your application and/or MPI
implementations. Some compilers may ha
Sofia,
I took your program and actually ran it successfully on my systems using
Open MPI r19400. A couple questions:
1. Have you tried to run the program on a single node?
mpirun -np 2 --host 10.4.5.123 --prefix /usr/local
./PruebaSumaParalela.out
2. Can you try and run the code the
Date: Wed, 17 Sep 2008 16:23:59 +0200
From: "Sofia Aparicio Secanellas"
Subject: Re: [OMPI users] Problem with MPI_Send and MPI_Recv
To: "Open MPI Users"
Message-ID: <0625EEFB84E04647A1930A963A8DF7E3@aparicio1>
Content-Type: text/plain; format=flowed; charset="iso-8859-1";
reply-type=r
It might also be interesting to see the result of "ifconfig -a" on both
of your machines.
--td
Date: Thu, 18 Sep 2008 10:19:37 +0200
From: "Sofia Aparicio Secanellas"
Subject: Re: [OMPI users] Problem with MPI_Send and MPI_Recv
To: "Open MPI Users"
Message-ID:
Content-Type: text/plain; form
Turns out you debugged the mpirun I was actually wanting you to attach to your program, PruebaSumaParalela.out, on both nodes and dump each of their stacks.
Is there a reason why you are using 1.2.2 instead of 1.2.7 or something from
the 1.3 branch? I am wondering if maybe there is some sort
Hello Sofia,
Ok, so I really wanted the stack of when you run with "-mca
mpi_preconnect_all 1" I believe you'll see that one of the processes
will be in init. However, the stack still probably will not help me
help you. What needs to happen is to step through the code in dbx while
the conn
Hello Sofia,
After further reflection I wonder if you have a firewall that is
preventing connections to certain ports.
--td
Terry Dontje wrote:
Hello Sofia,
Ok, so I really wanted the stack of when you run with "-mca
mpi_preconnect_all 1" I believe you'll see that one o
ocal -mca btl self,tcp -mca btl_tcp_if_include eth1
./PruebaSumaParalela.out I enclose you the results. Thank you. Sofia
- Original Message - From: "Terry Dontje"
To: Sent: Friday, September 19, 2008 7:54 PM
Subject: Re: [OMPI users] Problem with MPI_Send and MPI_Recv
> Hello Sofia,
>
Hello Sofia,
After talking with another OMPI member can you humor me and do
"/sbin/iptables -L" on both your machines. You'll need to be root to
do such.
--td
List-Post: users@lists.open-mpi.org
Date: Tue, 23 Sep 2008 06:02:30 -0400
From: Terry Dontje
Subject: Re: [OMPI user
Hello Sofia,
Very puzzling indeed. Can your try to run hostname or uptime with
mpirun? That is something like:
mpirun -np 2 --host 10.1.10.208,10.1.10.240 --mca mpi_preconnect_all 1
--prefix /usr/local -mca btl self,tcp -mca btl_tcp_if_include eth1 hostname
--td
List-Post: users@lists.o
Yann,
I'll take a look at this it looks like there definitely is an issue between our
libmpi.so and libmpi_f90.so files.
I noticed that the linkage message is a warning does the code actually fail
when running?
--td
List-Post: users@lists.open-mpi.org
Date: Tue, 07 Oct 2008 16:55:14 +0200
Fr
Yann,
How were you trying to link your code with PETSc? Did you use mpif90 or mpif77 wrappers or were you using cc or mpicc wrappers? I ran some basic tests that test the usage of MPI_STATUS_IGNORE using mpif90 (and mpif77) and it works fine. However I was able to generate a similar error as y
t/SUNWhpc/HPC8.0/lib/amd64/libmpi.so value=0x8; file
/opt/SUNWhpc/HPC8.0/lib/amd64/libmpi_f90.so value=0x14);
/opt/SUNWhpc/HPC8.0/lib/amd64/libmpi.so definition taken /usr/bin/rm
-f solv_ksp.o Thanks for your help, Yann Terry Dontje wrote:
> Yann,
>
> How were you trying to link yo
Yann,
Well, when you use f90 to link it passed the linker the -t option which is
described in the manpage with the following:
Turns off the warning for multiply-defined symbols that
have different sizes or different alignments.
That's why :-)
To your original question should y
We've found on certain applications binding to processors can have up to
a 2x difference. ScaliMPI automatically binds processes by socket so if
you are not running a one process per cpu job each process will land on
a different socket.
OMPI defaults to not binding at all. You may want to tr
A comment to the below. I meant the 2x performance was for shared
memory communications.
--td
Message: 3
Date: Wed, 05 Aug 2009 09:55:42 -0400
From: Terry Dontje
Subject: Re: [OMPI users] Performance difference on OpenMPI, IntelMPI
and ScaliMPI
To: us...@open-mpi.org
Message-ID
Craig,
Did your affinity script bind the processes per socket or linearly to cores.
If the former you'll want to look at using rankfiles and place the ranks based
on sockets. TWe have found this especially useful if you are not running fully
subscribed on your machines.
Also, if you think t
Hi Neeraj,
Were there specific collectives that were slower? Also what kind of cluster
were you running on? How many nodes and cores per node?
thanks,
--td
Message: 3
Date: Fri, 7 Aug 2009 16:51:05 +0530
From: nee...@crlindia.com
Subject: Re: [OMPI users] Performance question about OpenMPI
Date: Fri, 07 Aug 2009 07:12:45 -0600
From: Craig Tierney
Subject: Re: [OMPI users] Performance question about OpenMPI and
MVAPICH2 on IB
To: Open MPI Users
Message-ID: <4a7c284d.3040...@noaa.gov>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Terry Dontje
A copy of the configure line for Open MPI would be helpful. Which Intel
compiler are you using, version and bitness. Can you do file on
libmpi_f77.so? Also, are you sure that /usr/local/lib is where you
installed you Open MPI build and that isn't something latent?
--td
Date: Fri, 30 Oct 2
Georg,
I think your problem is you are using a ia32 (32 bit compiler) with a 64
bit built library. Either you need to use the intel64 compiler or build
Open MPI with the 32 bit library.
--td
*Subject:* Re: [OMPI users] Fortran Library Problem using openMPI
*From:* Georg A. Reichstein (/rei
Let me try this one more time.
Your application is being built with a 32 bit compiler ia32. However
the Open MPI libraries look to be built with the 64 bit compiler
intel64. One or the other needs to change.
--td
Terry Dontje wrote:
Georg,
I think your problem is you are using a ia32
Also, is the configure line you are giving below the application
configure line. I was actually asking for the Open MPI configure line.
--td
Terry Dontje wrote:
Let me try this one more time.
Your application is being built with a 32 bit compiler ia32. However
the Open MPI libraries look
Hi Basant,
I am not familiar with Windows builds of Open MPI. However, can you see
if you Open MPI build actually created a mca_paffinity_window's dll? I
could imagine the issue might be that the dll is not finding a needed
dependency. Under Windows is there a command similar to Unix's ldd
Technically MPI Spec may not put a requirement on TCP/IP, however Open
MPI's runtime environment needs some way to launch jobs and pass data
around in a standard way and it currently uses TCP/IP. That being said
there have been rumblings for some time to use other protocols but that
has not ye
A couple things to note. First Sun MPI 8.2.1 is effectively OMPI
1.3.4. I also reproduced the below issue using a C code so I think this
is a general issue with OMPI and not Fortran based.
I did a pmap of a process and there were two anon spaces equal to the
stack space set by ulimit.
In o
Date: Thu, 10 Dec 2009 17:57:27 -0500
From: Jeff Squyres
On Dec 10, 2009, at 5:53 PM, Gus Correa wrote:
> How does the efficiency of loopback
> (let's say, over TCP and over IB) compare with "sm"?
Definitely not as good; that's why we have sm. :-) I don't have any
quantificatio
Hi DevL, what compiler and options are you using to build OMPI. I am
seeing something similar (Warning messages and the Message Queue window
having bizarre values) when building with the Pathscale compiler but I
don't see this with SunStudio, gcc, Intel or PGI.
However, I do see pending recei
e the debugging
symbols.
Unfortunately, I still haven't tracked down Ashley's issue which I think
probably has more to do with the OMPI code instead of the debugging
information not being generated.
--td
Terry Dontje wrote:
Hi DevL, what compiler and options are you using to build OMPI
We haven't tried Solaris 8 in quite some time. However, for your first
issue did you include the --enable-heterogeneous option on your
configure command?
Since you are mix IA-32 and SPARC nodes you'll want to include this so
the endian issue doesn't bite you.
--td
Message: 5
Date: Thu, 04
Date: Fri, 05 Feb 2010 16:16:29 -0800 From: "David Mathog"
> We haven't tried Solaris 8 in quite some time. However, for your first
> issue did you include the --enable-heterogeneous option on your
> configure command?
>
> Since you are mix IA-32 and SPARC nodes you'll want to include this s
Jeff Squyres wrote:
Iain did the genius for the new assembly. Iain -- can you respond?
Iain is on vacation right now so he probably want be able to respond
until next week.
--td
On Feb 9, 2010, at 5:44 PM, Mostyn Lewis wrote:
The old opal_atomic_cmpset_32 worked:
static inline int
We (Oracle) have not done that much extensive limits testing going
between 32 to 64bit applications. Most of the testing we've done is
more around endianess (SPARC vs x86_64).
Though the below is kind of interesting. Sounds like the eager limit
isn't being normalized on the 64 bit machines.
FWIW, I took your code compiled it on a linux system using OMPI 1.4
r22761 and Solaris Studio C compilers. Then I ran it with "mpirun -np 4
a.out" and it seems to work for me:
Hello MPI World From process 0: Num processes: 4
Hello MPI World from process 1!
Hello MPI World from process 2!
Hello
Sorry, reposting this under the correct subject.
FWIW, I took your code compiled it on a linux system using OMPI 1.4
r22761 and Solaris Studio C compilers. Then I ran it with "mpirun -np 4
a.out" and it seems to work for me:
Hello MPI World From process 0: Num processes: 4
Hello MPI World from p
Hi Jonathan,
Do you know what the top level function is or communication pattern? Is
it some type of collective or a pattern that has a many to one. What
might be happening is that since OMPI uses a lazy connections by default
if all processes are trying to establish communications to the same
This looks like you are using an mpirun or mpiexec from mvapich to run
an executable compiled with OMPI. Can you make sure that you are using
the right mpirun?
--td
Pankatz, Klaus wrote:
Yes, I did that.
It ist basically the same problem with a Fortran version of this little
program. With
rs/pankatz/OPENmpi/bin/mpirun which is the right
one.
Von: users-boun...@open-mpi.org [users-boun...@open-mpi.org] im Auftrag von
Terry Dontje [terry.don...@oracle.com]
Gesendet: Freitag, 23. April 2010 14:29
An: Open MPI Users
Betreff: Re: [OMPI users] mp
How does the stack for the non-SM BTL run look, I assume it probably is
the same? Also, can you dump the message queues for rank 1? What's
interesting is you have a bunch of pending receives, do you expect that
to be the case when the MPI_Gatherv occurred?
--td
Teng Lin wrote:
Hi,
We rece
Does the "NODE 0: no room for lattice" message have any significance?
Could it be that for whatever reason the application is running out of
memory?
--td
kishore kumar wrote:
Hi,
I am trying to run SPEC MPI 2007 workload on a quad-core machine.
However getting this error message. I also tri
NGUYEN Laurent wrote:
Hello,
I'm trying to understand the difference between theses two options:
" --mca mpi_paffinity_alone 1 "
and
" --bind-to-core "
To me, it's the same thing (may be paffinity have maffinity in
addition) but the purpose af theses options is to bind MPI process to
process
NGUYEN Laurent wrote:
Ok, thank you for your answer. I think this rankfile feature is very
interesting to run some jobs like MPMD jobs or hybrid jobs
(multithreaded or GPU for examples).
Regards,
Point taken. The basic premise Jeff and I are working on is to see if
we could come up with a s
Gijsbert Wiesenekker wrote:
On May 11, 2010, at 9:29 , Gabriele Fatigati wrote:
Dear Gijsbert,
>Ideally I would like to check how many MPI_Isend messages have not
been processed yet, so that I can stop >sending messages if there are
'too many' waiting. Is there a way to do this?
you can
Olivier Riff wrote:
Hello,
I assume this question has been already discussed many times, but I
can not find on Internet a solution to my problem.
It is about buffer size limit of MPI_Send and MPI_Recv with
heterogeneous system (32 bit laptop / 64 bit cluster).
My configuration is :
open mpi 1
Which version of OMPI are you running on and the OS version?
Can you try and replace the rankfile specification with --bind-to-core
and tell me if that works any better?
--td
Chamila Janath wrote:
_rankfile_
rank 0=10.16.71.1 slot=0
I launched my mpi app using,
$ mpirun -np 1 -rf rankfile
It looks like the rankfile "*" syntax was broke between version r22761
and r23214. So, it looks like a regression to me. Ethan is looking
into trying to narrow this down more.
--td
Ralph Castain wrote:
I would have to look at the code, but I suspect it doesn't handle "*". Could be
upgraded
Sorry, there was a miscommunications between Ethan and I. The "*"
nomenclature never worked in OMPI, it is the specification of "n:*" that
works and we believe still works.
--td
Terry Dontje wrote:
It looks like the rankfile "*" syntax was broke between versi
Lydia, sorry I led you astray I meant for you to use the rankfile
feature as described in the mpirun manpage under the heading "Specifying
Ranks".
--td
Message: 1
Date: Mon, 5 Jan 2009 17:09:41 + (GMT)
From: Lydia Heck
Subject: [OMPI users] using the carto facility
To: us...@open-mpi.org
You know this all looks very similar to the reason why rolfv putback
r20351 which essentially defined out restrict within
opal_config_bottom.h when using Sun Studio.
--td
List-Post: users@lists.open-mpi.org
Date: Fri, 13 Mar 2009 16:40:49 -0400
From: Jeff Squyres
Subject: Re: [OMPI users] PGI
Sorry for the delay in response, I was out of the office late last week.
Can you tell me what version of Open MPI you are trying to build (1.2 or 1.3
branch)? Are you using the tarball on the Open MPI site or code downloaded
from the svn repository? Can you tell me which distribution and vers
I also was unable to reproduce the configure error with the latest 1.3
tarball. I was on a SLES distribution. What distribution are you on
and can you possibly try and configure using gcc instead of Sun Studio?
I have a feeling this issue is a larger configure issue and not Sun
Studio specif
Terry Dontje wrote:
I also was unable to reproduce the configure error with the latest 1.3
tarball. I was on a SLES distribution. What distribution are you on
and can you possibly try and configure using gcc instead of Sun
Studio? I have a feeling this issue is a larger configure issue and
Date: Mon, 30 Mar 2009 19:05:25 +0100
From: Kevin McManus
Subject: Re: [OMPI users] Linux opteron infiniband sunstudioconfigure
problem
To: Open MPI Users
Message-ID: <20090330180524.gt13...@gre.ac.uk>
Content-Type: text/plain; charset=us-ascii
> > I will try to reproduce the
ype; you must specify one
Can you manually run UNAME_REL=`(/bin/uname -X|grep Release|sed -e
's/.*= //')` in your shell without error?
--td
Terry Dontje wrote:
Date: Mon, 30 Mar 2009 19:05:25 +0100
From: Kevin McManus
Subject: Re: [OMPI users] Linux opteron infiniband sunstudio
c
Which version of the Sun Studio compilers are you using also which
version of OMPI are you trying to build. I am successful with building
OMPI with the Sun Studio Express release 200811 on Linux systems if I
don't use the C++ compiler. Prior releases we did suffer from some
issues. A "cc -V"
101 - 196 of 196 matches
Mail list logo
