Re: [O-MPI users] Problems running svn7132 with more than on proc per node

Thanks Peter, We'll look into this... Tim Peter Kjellström wrote: Hello, I'm playing with a copy of svn7132 that built and installed just fine. At first everything seemed ok, unlike earlier it now runs on mvapi automagically :-) But then a small testprogram failed and then another. After

Re: [O-MPI users] attempt to override read-only MCA parameter "orte_infrastructure"

Hello Daryl, I believe there is a problem w/ the latest version of the bproc launcher... Try running w/ the following to use an older version: mpirun -mca pls_bproc_seed_priority 101 This could also be set in your system default or local MCA parameter file. Thanks, Tim Daryl W. Grunau wrot

Re: [O-MPI users] attempt to override read-only MCA parameter "orte_infrastructure"

Daryl, I'm seeing the same error messages - we'll look into this. Does the mvapi port appear to be working for you. W/ a debug build I'm seeing: Latency: mpi-ping: ping-pong nprocs=2, reps=1, min bytes=0, max bytes=0 inc bytes=0 0 pings 1 0 pinged 1:0 bytes 7.47 uSec 0.

Re: [O-MPI users] attempt to override read-only MCA parameter "orte_infrastructure"

Daryl, This should be fixed in svn. Thanks, Tim Daryl W. Grunau wrote: Hi, I downloaded/installed version 1.0a1r7337 configured to run on my BProcV4 IB cluster (mvapi, for now). Upon execution, I get the following warning message, however the app appears to run to completion afterwards:

Re: [O-MPI users] attempt to override read-only MCA parameter "orte_infrastructure" - FIXED!

Daryl, Tim, the latest nightly fixes this - thanks! Can I report another? I can't seem to specify -H|-host|--host ; mpirun seems to ignore the argument: % mpirun -np 2 -H 0,4 ./cpi Process 0 on n0 Process 1 on n1 pi is approximately 3.1416009869231241, Error is 0.08

Re: [O-MPI users] attempt to override read-only MCA parameter "orte_infrastructure" - FIXED!

r the time being. On Sep 19, 2005, at 6:12 PM, Tim S. Woodall wrote: Daryl, Tim, the latest nightly fixes this - thanks! Can I report another? I can't seem to specify -H|-host|--host ; mpirun seems to ignore the argument: % mpirun -np 2 -H 0,4 ./cpi Process 0 on n0 Proce

Re: [O-MPI users] HOWTO turn of "multi-rail" support at runtime?

Daryl, Try setting: -mca btl_base_include self,mvapi To specify that only lookback (self) and mvapi btls should be used. Can you forward me the config.log from your build? Thanks, Tim Daryl W. Grunau wrote: Hi, I've got a dual-homed IB + GigE connected cluster for which I've built a very re

Re: [O-MPI users] OpenMPI hang issue

Hello Chris, Please give the next release candidate a try. There was an issue w/ the GM port that was likely causing this. Thanks, Tim Parrott, Chris wrote: Greetings, I have been testing OpenMPI 1.0rc3 on a rack of 8 2-processor (single core) Opteron systems connected via both Gigabit Ether

Re: [O-MPI users] Hpl Bench mark and Openmpi rc3 (Jeff Squyres)

Hi Jeff, I installed two versions of open mpi slightly different. One on /opt/openmpi or I would get the gfortran error and the other in /home/allan/openmpi However I do not think that is the problem as the path names are specified in the bahrc and bash_profile files of the /home/allan dire

Re: [O-MPI users] TCP problems with 1.0rc4

Mike, Mike Houston wrote: We can't seem to run across TCP. We did a default 'configure'. Shared memory seems to work, but trying tcp give us: [0,1,1][btl_tcp_endpoint.c:557:mca_btl_tcp_endpoint_complete_connect] connect() failed with errno=113 This error indicates the IP address exporte

Re: [O-MPI users] TCP problems with 1.0rc4

This error indicates the IP address exported by the peer is not reachable. You can use the tcp btl parameters: -mca btl_tcp_include eth0,eth1 or -mca btl_tcp_exclude eth1 To specify the set of interfaces to use/not use. George was correct - these should be btl_tcp_if_include/btl_tcp_if_

Re: [O-MPI users] Infiniband performance problems (mvapi)

Hello Mike, Mike Houston wrote: When only sending a few messages, we get reasonably good IB performance, ~500MB/s (MVAPICH is 850MB/s). However, if I crank the number of messages up, we drop to 3MB/s(!!!). This is with the OSU NBCL mpi_bandwidth test. We are running Mellanox IB Gold 1.8 wit

Re: [O-MPI users] Infiniband performance problems (mvapi)

p performance. Now if I can get the tcp layer working, I'm pretty much good to go. Any word on an SDP layer? I can probably modify the tcp layer quickly to do SDP, but I thought I would ask. -Mike Tim S. Woodall wrote: Hello Mike, Mike Houston wrote: When only sending a fe

Re: [O-MPI users] Infiniband performance problems (mvapi)

Mike, Let me confirm this was the issue and look at the TCP problem as well. Will let you know. Thanks, Tim Mike Houston wrote: What's the ETA, or should I try grabbing from cvs? -Mike Tim S. Woodall wrote: Mike, I believe was probably corrected today and should be in the next re

Re: [O-MPI users] Infiniband performance problems (mvapi)

r should I try grabbing from cvs? -Mike Tim S. Woodall wrote: Mike, I believe was probably corrected today and should be in the next release candidate. Thanks, Tim Mike Houston wrote: Woops, spoke to soon. The performance quoted was not actually going between nodes. Actually using the

Re: [O-MPI users] Infiniband performance problems (mvapi)

Any word on an SDP layer? I can probably modify the tcp layer quickly to do SDP, but I thought I would ask. -Mike Tim S. Woodall wrote: Hello Mike, Mike Houston wrote: When only sending a few messages, we get reasonably good IB performance, ~500MB/s (MVAPICH is 850MB/s). H

Re: [O-MPI users] problems with bproc cluster

Hello John, You need to specify both --enable-static and --disable-shared to do a static build (not sure why, perhaps someone else can fill us in on that)... The logs indicate the launch is failing trying to start orted on the backend node... probably due to shared library dependencies. You mig

Re: [O-MPI users] problems with bproc cluster

John, Any progress on this? John Ouellette wrote: Hi Tim, H, nope. I recompiled OpenMPI to produce the static libs, and even recompiled my app statically, and received the same error messages. If orted isn't starting on the compute nodes, is there any way I can debug this to find out

Re: [O-MPI users] error creating high priority cq for mthca0

Daryl, Try this: Original Message Subject: RE: only root running mpi jobs with 1.0.1rc5 List-Post: users@lists.open-mpi.org Date: Thu, 01 Dec 2005 18:49:46 -0700 From: Joshua Aune Reply-To: lu...@lnxi.com Organization: Linux Networx To: Todd Wilde CC: Matthew Finlay , twood.

Re: [OMPI users] Memory allocation problem with OpenIB

Hello Emanuel, You might want to try an actual hard limit, say 8GB, rather than unlimited. I've run into issues w/ unlimited in the past. Thanks, Tim Emanuel Ziegler wrote: Hi! After solving my last problem with the help of this list (thanks again :) I encountered another problem regarding t

Re: [OMPI users] Building 32-bit OpenMPI package for 64-bit Opteron platform

Ralph/all, Ralph Castain wrote: Unfortunately, that's all that is available at the moment. Future releases (post 1.1) may get around this problem. The issue is that the bproc launcher actually does a binary memory image of the process, then replicates that across all the nodes. This is how w