Anyone can help me to find out problem or bug in my cluster? The output of
"ibv_devinfo -v" from Dell blade with infiniband module look very strange. The
phys_port_cnt is 2, one active, and another down. The active port is 20x speed,
the down port is 10x speed. We are using Dell PowerEdge M600 B
Excellent; I'm glad that these FAQ items are useful!
On May 7, 2009, at 10:03 AM, > wrote:
I found a prior message from Jeff Squyres that answered my
question. His
message gave a link to the FAQ which gave me what I need.
Thanks,
Pat
Link for configurati
Sorry about the typo, yes, I meant OMPI 1.3.2.
Mehdi
-Original Message-
From: users-boun...@open-mpi.org [mailto:users-boun...@open-mpi.org] On
Behalf Of Jeff Squyres
Sent: May-07-09 12:07 PM
To: Open MPI Users
Subject: Re: [OMPI users] LSF launch with OpenMPI
Did you mean OMPI 1.3.2?
O
Did you mean OMPI 1.3.2?
OMPI 1.2.3 did not have LSF support.
On May 7, 2009, at 9:50 AM, Mehdi Bozzo-Rey wrote:
Hi Jeff,
I tried several combinations and:
- LIBS=... does not work for OpenMPI 1.2.3 / LSF 7.0.5
- the winner for now is LSF 7.0.4 / OpenMPI 1.2.3
Cheers,
Mehdi
-Original
Hi,
Thank you for the tip, this seems to be what I was looking for.
Matthieu
2009/5/7 Mehdi Bozzo-Rey :
> Hello Jeroen,
>
>
>
> There are 2 ways of launching OpenMPI jobs (using a recent version of LSF):
>
> 1. The one you have just described; it uses the generic PJL (Parallel
> Job Launch
I found a prior message from Jeff Squyres that answered my question. His
message gave a link to the FAQ which gave me what I need.
Thanks,
Pat
Link for configuration:
http://www.open-mpi.org/faq/?category=building#default-build
J.W. (Pat) O'Bryant,Jr.
Business
The (low level verbs) latency has AFAIR changed only a few times:
1) started at 5-6us with PCI-X Infinihost3
2) dropped to 3-4us with PCI-express Infinihost3
3) dropped to ~1us with PCI-express ConnectX
I would like to add that on PCI-EX Gen2 platforms the latency is sub
micro (~0.8-0.95)
I am in the process of building a production system with OpenMPI 1.3.2
with support for OFED. Is it necessary in the "configure" statement to
specify "--with-openib(=DIR)" to get OFED support? I have built a test
system with OpenMPI 1.3.2 and an "ompi_info" yields the output below. It
appears
Hi Jeff,
I tried several combinations and:
- LIBS=... does not work for OpenMPI 1.2.3 / LSF 7.0.5
- the winner for now is LSF 7.0.4 / OpenMPI 1.2.3
Cheers,
Mehdi
-Original Message-
From: users-boun...@open-mpi.org [mailto:users-boun...@open-mpi.org] On
Behalf Of Jeff Squyres
Sent: May-
Hi Jeff,
I just tried it: OpenMPI 1.3.2 (compiled with no LSF support)/ LSF 7.0.4
and the PJL framework (-a openmpi / mpirun.lsf) and everything looks
fine.
Cheers,
Mehdi
-Original Message-
From: users-boun...@open-mpi.org [mailto:users-boun...@open-mpi.org] On
Behalf Of Jeff Squyres
Se
Hello Jeroen,
There are 2 ways of launching OpenMPI jobs (using a recent version of
LSF):
1. The one you have just described; it uses the generic PJL
(Parallel Job Launcher) framework. You can easily recognise it because
of the use of the -a openmpi flag and mpirun.lsf
2. In recen
Jeff Squyres wrote:
On May 6, 2009, at 4:45 PM, Ken Cain wrote:
Is it possible for OMPI to generate output at runtime indicating exactly
what btl(s) will be used?
At present, we only have a fairly lame system to do this. We wanted to
print out a connection map in v1.3, but it didn't happen
On May 6, 2009, at 4:45 PM, Ken Cain wrote:
Is it possible for OMPI to generate output at runtime indicating
exactly
what btl(s) will be used?
At present, we only have a fairly lame system to do this. We wanted
to print out a connection map in v1.3, but it didn't happen -- this
feature
I'm glad that the recent commits fixed your problem.
At the moment, we do not implement a mirroring file storage mechanism
(where peers save checkpoints to each others local disk). We have
been working towards supporting this and other techniques in some off-
trunk development, but nothing r
I just did, at least the ping pong, the results are slightly worst and
presents the same drop at 64KSee attachment.
a comment: to run with the btl mx i need to use --mca btl mx,sm,self -mca
mtl ^mx or i get a mx_open_endpoint failure due to myrinet busy (I have
already increase the number of en
On Thursday 07 May 2009, nee...@crlindia.com wrote:
> Thanks Pasha for sharing IB Roadmaps with us. But i am more interested in
> to find out latency figures since they often matter more than bit rate.
>
> Could there be rough if not accurate the latency figures being targeted in
> IB World?
The (
Hello,
Thank you, with the release r21172 and it works. But how i can dispatch
the checkpoint on different storage nodes, because it is to costly that
all computing nodes write on one storage node.
Josh Hursey a écrit :
I just realized that not all of the FileM fixes made it to the trunk
i
Thanks Pasha for sharing IB Roadmaps with us. But i am more interested in
to find out latency figures since they often matter more than bit rate.
Could there be rough if not accurate the latency figures being targeted in
IB World?
Regards
Neeraj Chourasia
Member of Technical Staff
Computation
18 matches
Mail list logo
