<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
<style type="text/css" style="display:none;"> P {margin-top:0;margin-bottom:0;} </style>
</head>
<body dir="ltr">
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
FYI, I opened a GitLab issue to track this problem: <a href="https://xgitlab.cels.anl.gov/darshan/darshan/issues/270" id="LPlnk317212">https://xgitlab.cels.anl.gov/darshan/darshan/issues/270</a><br>
</div>
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
The crashes are a new surprise to us, so would like to have a look before our next release. I'll look to see if I can get a newer Lustre version working in a container or something to test against and will report back what I find.<br>
</div>
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
<br>
</div>
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
Thanks,</div>
<div style="font-family: Calibri, Arial, Helvetica, sans-serif; font-size: 12pt; color: rgb(0, 0, 0);">
--Shane<br>
</div>
<div id="appendonsend"></div>
<hr style="display:inline-block;width:98%" tabindex="-1">
<div id="divRplyFwdMsg" dir="ltr"><font face="Calibri, sans-serif" style="font-size:11pt" color="#000000"><b>From:</b> Darshan-users <darshan-users-bounces@lists.mcs.anl.gov> on behalf of Carns, Philip H. <carns@mcs.anl.gov><br>
<b>Sent:</b> Thursday, April 2, 2020 10:27 AM<br>
<b>To:</b> André R. Carneiro <andre.es@gmail.com>; darshan-users@lists.mcs.anl.gov <darshan-users@lists.mcs.anl.gov><br>
<b>Subject:</b> Re: [Darshan-users] Darshan crashing compute node</font>
<div> </div>
</div>
<style type="text/css" style="display:none">
<!--
p
{margin-top:0;
margin-bottom:0}
-->
</style>
<div dir="ltr">
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
Thanks for the bug reports André! You are describing two different problems:</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
<br>
</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
1) incompatibility between the Darshan Lustre instrumentation and recent Lustre releases: we'll definitely look into this. As you found in the mailing list archives there has been a known problem here, but this is a new permutation that it causes a node crash
now. That's disappointing, to say the least. If anyone happens to have a pointer to how to stand up a Lustre 2.11 instance in docker or a VM they can share, that would be greatly appreciated and would help us track compatibility better. The Lustre systems
we have access to are not new enough. In the mean time you are correct to disable that component at build time.</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
<br>
</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
2) floating point or divide by zero when writing log to Lustre: This one is odd because at the point of the problem, Darshan is not instrumenting anything: it is simply an application writing data to Lustre. We don't generally see problems here. Darshan is
doing a collective write here as the last step of aggregating instrumentation from all ranks into a single log file. The only slightly unusual things that Darshan could be doing are a) setting MPI-IO hints or b) using particular datatypes to organize the
writes. Neither should be a problem, but it looks like they might have triggered a bug somewhere.</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
<br>
</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
If you don't mind trying another test case, could you repeat one of the experiments that crashes with a floating point or divide by zero bug down in MPI_File_write_at_all() with the following environment variable set? (please keep disabling the Lustre module
at build time, we have no other fix for that ready yet) </div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
<br>
</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
export DARSHAN_LOGHINTS=</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
<br>
</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
(in other words, set the DARSHAN_LOGHINTS environment variable to an empty string)</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
<br>
</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
I believe that this will clear the MPI-IO hints that Darshan would normally specify when opening the output log file. If that makes your example work cleanly, then we can narrow down which hint is the problem and maybe get some help outside of Darshan for
the root cause. If it still fails, then we need to look elsewhere for the problem.</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
<br>
</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
thanks,</div>
<div style="font-family:Calibri,Arial,Helvetica,sans-serif; font-size:12pt; color:rgb(0,0,0); background-color:rgb(255,255,255)">
-Phil</div>
<div id="x_appendonsend"></div>
<hr tabindex="-1" style="display:inline-block; width:98%">
<div id="x_divRplyFwdMsg" dir="ltr"><font face="Calibri, sans-serif" color="#000000" style="font-size:11pt"><b>From:</b> Darshan-users <darshan-users-bounces@lists.mcs.anl.gov> on behalf of André R. Carneiro <andre.es@gmail.com><br>
<b>Sent:</b> Wednesday, April 1, 2020 1:39 PM<br>
<b>To:</b> darshan-users@lists.mcs.anl.gov <darshan-users@lists.mcs.anl.gov><br>
<b>Subject:</b> [Darshan-users] Darshan crashing compute node</font>
<div> </div>
</div>
<div>
<div dir="ltr">
<div>Hi,</div>
<div><br>
</div>
<div>I'm traying to use darshan (tested all 3.1.X versions) on a HPC cluster running with Red Hat 7.6 (kernel 3.10.0-957.el7.x86_64) and using Lustre Filesystem version 2.11.0.300_cray_102_g3dbace1, but when the application starts, the compute node crashes
with a kernel panic message on the console. </div>
<div><br>
</div>
<div>While searching on the Darshan mailing list I came across this message: <a href="https://lists.mcs.anl.gov/mailman/htdig/darshan-users/2019-October/000542.html">https://lists.mcs.anl.gov/mailman/htdig/darshan-users/2019-October/000542.html</a></div>
<div><br>
</div>
<div>I disabled the Lustre module and recompiled darshan. The crash isn't happening anymore, but now it's unable to write the darshan log file on the lustre filesystem, with the following error:</div>
<div><br>
</div>
<div>
<div>*Using OpenMPI 3.1.5 and GCC 7</div>
<div></div>
</div>
<div><br>
</div>
<div>Program received signal SIGFPE: Floating-point exception - erroneous arithmetic operation.<br>
Backtrace for this error:<br>
#0 0x7f4f6759f27f in ???<br>
#1 0x7f4f687ababe in ???<br>
#2 0x7f4f687add06 in ???<br>
#3 0x7f4f687db6c0 in ???<br>
#4 0x7f4f687dbddb in ???<br>
#5 0x7f4f6879d6f1 in ???<br>
#6 0x7f4f6871892b in ???<br>
#7 0x7f4f691d0ae1 in MPI_File_write_at_all<br>
at lib/darshan-mpiio.c:536<br>
#8 0x7f4f691bea7f in darshan_log_append_all<br>
at lib/darshan-core.c:1800<br>
#9 0x7f4f691c1907 in darshan_log_write_name_record_hash<br>
at lib/darshan-core.c:1761<br>
#10 0x7f4f691c1907 in darshan_core_shutdown<br>
at lib/darshan-core.c:546<br>
#11 0x7f4f691be402 in MPI_Finalize<br>
at lib/darshan-core-init-finalize.c:82<br>
#12 0x7f4f68b6a798 in ???<br>
#13 0x4023bb in ???<br>
#14 0x401ae6 in ???<br>
#15 0x7f4f6758b3d4 in ???<br>
#16 0x401b16 in ???<br>
#17 0xffffffffffffffff in ???<br>
--------------------------------------------------------------------------<br>
</div>
<div><br>
</div>
<div>*Using Intel PSXE 2018 with Intel MPI</div>
<div><br>
</div>
<div>forrtl: severe (71): integer divide by zero<br>
Image PC Routine Line Source
<br>
exec.exe 000000000045282E Unknown Unknown Unknown<br>
libpthread-2.17.s 00002B8B5A5FE5D0 Unknown Unknown Unknown<br>
libmpi_lustre.so. 00002B8B659D4FDF ADIOI_LUSTRE_Get_ Unknown Unknown<br>
libmpi_lustre.so. 00002B8B659CFFD9 ADIOI_LUSTRE_Writ Unknown Unknown<br>
libmpi.so.12.0 00002B8B59A4C15C Unknown Unknown Unknown<br>
libmpi.so.12 00002B8B59A4D1D5 PMPI_File_write_a Unknown Unknown<br>
libdarshan.so 00002B8B58F90312 MPI_File_write_at Unknown Unknown<br>
libdarshan.so 00002B8B58F7E63A Unknown Unknown Unknown<br>
libdarshan.so 00002B8B58F815B0 darshan_core_shut Unknown Unknown<br>
libdarshan.so 00002B8B58F7DFF3 MPI_Finalize Unknown Unknown<br>
libmpifort.so.12. 00002B8B592414DA pmpi_finalize__ Unknown Unknown<br>
exec.exe 00000000004490A5 Unknown Unknown Unknown<br>
exec.exe 00000000004032DE Unknown Unknown Unknown<br>
<a href="http://libc-2.17.so">libc-2.17.so</a> 00002B8B5AB2F3D5 __libc_start_main Unknown Unknown<br>
exec.exe 00000000004031E9 Unknown Unknown Unknown<br>
</div>
<div><br>
</div>
<div><br>
</div>
<div>If I configure darshan to write on a different filesystem (local /tmp of the first compute node), it works out fine, but then I'm restricted to use only one compute node, since the output dir have to be shared among all nodes (MPI tasks). <br>
</div>
<div><br>
</div>
<div>Is there a workaround for this? At the moment, my cluster only have the lustre filesystem as a shared filesystem among all compute nodes.</div>
<div><br>
</div>
Best regards.<br clear="all">
<div><br>
</div>
-- <br>
<div dir="ltr" class="x_x_gmail_signature">Abraços³, <br>
André Ramos Carneiro.</div>
</div>
</div>
</div>
</body>
</html>