<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<p>Hello,<br>
<br>
Yes, we have recompiled OpenMPI with integration with SLURM 19.05
but the problem remains.<br>
<br>
We have also tried to recompile OpenMPI without integration with
SLURM. In this case executions fail with srun, but with mpirun it
continues to work in SLURM 18.08 and fails in 19.05 in the same
way.<br>
<br>
Thank you very much once more.</p>
<p><br>
</p>
<div class="moz-cite-prefix">El 06/06/2019 a las 12:46, Sean Crosby
escribió:<br>
</div>
<blockquote type="cite"
cite="mid:CADg9O+M2uQhYDYi9fM-DcYhVu_Eu6=jBzUafrJ4jr7m1NFANAg@mail.gmail.com">
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
<div dir="ltr">
<div>Hi Andrés,</div>
<div><br>
</div>
<div>Did you recompile OpenMPI after updating to SLURM 19.05?</div>
<div><br>
</div>
<div>Sean<br>
</div>
<div>
<div>
<div dir="ltr" class="gmail_signature"
data-smartmail="gmail_signature">
<div dir="ltr">
<div>
<div dir="ltr">
<div>
<div dir="ltr">
<div dir="ltr">
<div dir="ltr">
<div dir="ltr"><br>
</div>
<div dir="ltr">--</div>
<div dir="ltr">Sean Crosby</div>
<div dir="ltr">
<div dir="ltr">Senior DevOpsHPC Engineer
and HPC Team Lead | Research Platform
Services</div>
<div>Research Computing | CoEPP | School
of Physics<br>
</div>
</div>
<div dir="ltr"><span
style="font-size:12.8px">University of
Melbourne</span><br>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
<br>
</div>
</div>
<br>
<div class="gmail_quote">
<div dir="ltr" class="gmail_attr">On Thu, 6 Jun 2019 at 20:11,
Andrés Marín Díaz <<a href="mailto:amarin@cesvima.upm.es"
moz-do-not-send="true">amarin@cesvima.upm.es</a>> wrote:<br>
</div>
<blockquote class="gmail_quote" style="margin:0px 0px 0px
0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
Thank you very much for the help, I update some information.<br>
<br>
- If we use Intel MPI (IMPI) mpirun it works correctly.<br>
- If we use mpirun without using the scheduler it works
correctly.<br>
- If we use srun with software compiled with OpenMPI it works
correctly.<br>
- If we use SLURM 18.08.6 it works correctly.<br>
- If we use SLURM 19.05.0 and mpirun inside the sbatch script
then we <br>
get the error:<br>
--------------------------------------------------------------------------<br>
An ORTE daemon has unexpectedly failed after launch and
before<br>
communicating back to mpirun. This could be caused by a
number<br>
of factors, including an inability to create a connection
back<br>
to mpirun due to lack of common network interfaces and /
or no<br>
route found between them. Please check network
connectivity<br>
(including firewalls and network routing requirements).<br>
--------------------------------------------------------------------------<br>
<br>
Trying to trace the problem:<br>
- Mpirun is a binary and can not be traced with batch -x.<br>
- I've done a "strace mpirun hostname" to see if it helps, but
i am not <br>
able to see where the problem may be.<br>
<br>
Here is the exit from the strace:<br>
<a
href="https://cloud.cesvima.upm.es/index.php/s/hWQMkwU5zW7J8RW"
rel="noreferrer" target="_blank" moz-do-not-send="true">https://cloud.cesvima.upm.es/index.php/s/hWQMkwU5zW7J8RW</a><br>
<br>
And here the slurmd log with verbose level 5:<br>
Main node (slurmd log):<br>
2019-06-06T09:51:54.255743+00:00 r1n1 slurmd[108517]:
_run_prolog: <br>
run job script took usec=7<br>
2019-06-06T09:51:54.256118+00:00 r1n1 slurmd[108517]:
_run_prolog: <br>
prolog with lock for job 11057 ran for 0 seconds<br>
2019-06-06T09:51:54.258887+00:00 r1n1 slurmd[108517]: <br>
task_p_slurmd_batch_request: 11057<br>
2019-06-06T09:51:54.259317+00:00 r1n1 slurmd[108517]: <br>
task/affinity: job 11057 CPU input mask for node: 0x0000000001<br>
2019-06-06T09:51:54.259680+00:00 r1n1 slurmd[108517]: <br>
task/affinity: job 11057 CPU final HW mask for node:
0x0000000001<br>
2019-06-06T09:51:54.279614+00:00 r1n1 slurmstepd[108548]:
task <br>
affinity plugin loaded with CPU mask <br>
000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000ffffffffff<br>
2019-06-06T09:51:54.280312+00:00 r1n1 slurmstepd[108548]:
Munge <br>
credential signature plugin loaded<br>
2019-06-06T09:51:54.302921+00:00 r1n1 slurmstepd[108548]:
<br>
task/cgroup: /slurm/uid_2000/job_11057: alloc=1024MB
mem.limit=1024MB <br>
memsw.limit=unlimited<br>
2019-06-06T09:51:54.303592+00:00 r1n1 slurmstepd[108548]:
<br>
task/cgroup: /slurm/uid_2000/job_11057/step_extern:
alloc=1024MB <br>
mem.limit=1024MB memsw.limit=unlimited<br>
2019-06-06T09:51:54.330322+00:00 r1n1 slurmd[108517]:
Launching <br>
batch job 11057 for UID 2000<br>
2019-06-06T09:51:54.353196+00:00 r1n1 slurmstepd[108556]:
task <br>
affinity plugin loaded with CPU mask <br>
000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000ffffffffff<br>
2019-06-06T09:51:54.353899+00:00 r1n1 slurmstepd[108556]:
Munge <br>
credential signature plugin loaded<br>
2019-06-06T09:51:54.366478+00:00 r1n1 slurmstepd[108556]:
<br>
task/cgroup: /slurm/uid_2000/job_11057: alloc=1024MB
mem.limit=1024MB <br>
memsw.limit=unlimited<br>
2019-06-06T09:51:54.366755+00:00 r1n1 slurmstepd[108556]:
<br>
task/cgroup: /slurm/uid_2000/job_11057/step_batch:
alloc=1024MB <br>
mem.limit=1024MB memsw.limit=unlimited<br>
2019-06-06T09:51:54.393325+00:00 r1n1 slurmstepd[108556]:
debug <br>
level = 2<br>
2019-06-06T09:51:54.393754+00:00 r1n1 slurmstepd[108556]:
starting <br>
1 tasks<br>
2019-06-06T09:51:54.401243+00:00 r1n1 slurmstepd[108556]:
task 0 <br>
(108561) started 2019-06-06T09:51:54<br>
2019-06-06T09:51:54.416396+00:00 r1n1 slurmstepd[108561]:
<br>
task_p_pre_launch: Using sched_affinity for tasks<br>
2019-06-06T09:51:56.514908+00:00 r1n1 slurmstepd[108556]:
task 0 <br>
(108561) exited with exit code 1.<br>
2019-06-06T09:51:56.554430+00:00 r1n1 slurmstepd[108556]:
job 11057 <br>
completed with slurm_rc = 0, job_rc = 256<br>
2019-06-06T09:51:56.554847+00:00 r1n1 slurmstepd[108556]:
sending <br>
REQUEST_COMPLETE_BATCH_SCRIPT, error:0 status:256<br>
2019-06-06T09:51:56.559856+00:00 r1n1 slurmstepd[108556]:
done with job<br>
2019-06-06T09:51:56.596762+00:00 r1n1 slurmstepd[108548]:
Sent <br>
signal 18 to 11057.4294967295<br>
2019-06-06T09:51:56.598072+00:00 r1n1 slurmstepd[108548]:
Sent <br>
signal 15 to 11057.4294967295<br>
2019-06-06T09:51:56.599141+00:00 r1n1 slurmstepd[108548]:
<br>
_oom_event_monitor: oom-kill event count: 1<br>
2019-06-06T09:51:56.641170+00:00 r1n1 slurmstepd[108548]:
done with job<br>
<br>
Secundary node (slurmd log):<br>
2019-06-06T09:51:54.256047+00:00 r1n2 slurmd[84916]:
_run_prolog: <br>
run job script took usec=7<br>
2019-06-06T09:51:54.256432+00:00 r1n2 slurmd[84916]:
_run_prolog: <br>
prolog with lock for job 11057 ran for 0 seconds<br>
2019-06-06T09:51:54.279763+00:00 r1n2 slurmstepd[84954]:
task <br>
affinity plugin loaded with CPU mask <br>
000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000ffffffffff<br>
2019-06-06T09:51:54.280448+00:00 r1n2 slurmstepd[84954]:
Munge <br>
credential signature plugin loaded<br>
2019-06-06T09:51:54.313852+00:00 r1n2 slurmstepd[84954]:
<br>
task/cgroup: /slurm/uid_2000/job_11057: alloc=1024MB
mem.limit=1024MB <br>
memsw.limit=unlimited<br>
2019-06-06T09:51:54.314502+00:00 r1n2 slurmstepd[84954]:
<br>
task/cgroup: /slurm/uid_2000/job_11057/step_extern:
alloc=1024MB <br>
mem.limit=1024MB memsw.limit=unlimited<br>
2019-06-06T09:51:56.564758+00:00 r1n2 slurmstepd[84954]:
Sent <br>
signal 18 to 11057.4294967295<br>
2019-06-06T09:51:56.608028+00:00 r1n2 slurmstepd[84954]:
Sent <br>
signal 15 to 11057.4294967295<br>
2019-06-06T09:51:56.609259+00:00 r1n2 slurmstepd[84954]:
<br>
_oom_event_monitor: oom-kill event count: 1<br>
2019-06-06T09:51:56.638334+00:00 r1n2 slurmstepd[84954]:
done with job<br>
<br>
Thank you very much again.<br>
<br>
-- <br>
Andrés Marín Díaz<br>
<br>
Servicio de Infraestructura e Innovación<br>
Universidad Politécnica de Madrid<br>
<br>
Centro de Supercomputación y Visualización de Madrid
(CeSViMa)<br>
Campus de Montegancedo. 28223, Pozuelo de Alarcón, Madrid
(ES)<br>
<a href="mailto:amarin@cesvima.upm.es" target="_blank"
moz-do-not-send="true">amarin@cesvima.upm.es</a> | tel
910679676<br>
<br>
<a href="http://www.cesvima.upm.es" rel="noreferrer"
target="_blank" moz-do-not-send="true">www.cesvima.upm.es</a>
|
<a href="http://www.twitter.com/cesvima" rel="noreferrer"
target="_blank" moz-do-not-send="true">www.twitter.com/cesvima</a>
|
<a href="http://www.fb.com/cesvima" rel="noreferrer"
target="_blank" moz-do-not-send="true">www.fb.com/cesvima</a><br>
<br>
<br>
<br>
</blockquote>
</div>
</blockquote>
<pre class="moz-signature" cols="72">--
Andrés Marín Díaz
Servicio de Infraestructura e Innovación
Universidad Politécnica de Madrid
Centro de Supercomputación y Visualización de Madrid (CeSViMa)
Campus de Montegancedo. 28223, Pozuelo de Alarcón, Madrid (ES)
<a class="moz-txt-link-abbreviated" href="mailto:amarin@cesvima.upm.es">amarin@cesvima.upm.es</a> | tel 910679676
<a class="moz-txt-link-abbreviated" href="http://www.cesvima.upm.es">www.cesvima.upm.es</a> | <a class="moz-txt-link-abbreviated" href="http://www.twitter.com/cesvima">www.twitter.com/cesvima</a> | <a class="moz-txt-link-abbreviated" href="http://www.fb.com/cesvima">www.fb.com/cesvima</a>
</pre>
</body>
</html>