Project

General

Profile

Bug #2016

+autoProvision ignores numactl settings

Added by Jim Phillips 19 days ago. Updated 19 days ago.

Status:
New
Priority:
Normal
Assignee:
Category:
Machine Layers
Target version:
-
Start date:
10/26/2018
Due date:
% Done:

0%


Description

Runs on 64 threads when numactl is limited to 12:

policy: default
preferred node: current
physcpubind: 0 1 2 3 4 5 6 7 8 9 10 11
cpubind: 0 1
nodebind: 0 1
membind: 0 1 2 3 4 5 6 7
Charm++: standalone mode (not using charmrun)
Charm++> Running in Multicore mode: 64 threads (PEs)
Charm++> Using recursive bisection (scheme 3) for topology aware partitions
Converse/Charm++ Commit ID: v6.9.0-rc2-0-ge8d6e0c-namd-charm-6.9.0-build-2018-Oct-25-11430

Note that it does follow Comet's queueing system setup, and warns that there is something odd:

Intel(R) processor family information utility, Version 4.1 Update 3 Build 20131205
Copyright (C) 2005-2013 Intel Corporation.  All rights reserved.

=====  Processor composition  =====
Processor name    : Intel(R) Xeon(R)  E5-2680 v3 
Packages(sockets) : 1
Cores             : 12
Processors(CPUs)  : 12
Cores per package : 12
Threads per core  : 1

=====  Processor identification  =====
Processor    Thread Id.    Core Id.    Package Id.
0           0           0           0   
1           0           1           0   
2           0           2           0   
3           0           3           0   
4           0           4           0   
5           0           5           0   
6           0           8           0   
7           0           9           0   
8           0           10          0   
9           0           11          0   
10          0           12          0   
11          0           13          0   
=====  Placement on packages  =====
Package Id.    Core Id.    Processors
0           0,1,2,3,4,5,8,9,10,11,12,13        0,1,2,3,4,5,6,7,8,9,10,11

=====  Cache sharing  =====
Cache    Size        Processors
L1    32  KB        no sharing
L2    256 KB        no sharing
L3    30  MB        (0,1,2,3,4,5,6,7,8,9,10,11)
policy: default
preferred node: current
physcpubind: 0 1 2 3 4 5 6 7 8 9 10 11 
cpubind: 0 
nodebind: 0 
membind: 0 1 
/var/spool/slurmd/job19830854/slurm_script: line 14: nvidia-smi: command not found
Charm++: standalone mode (not using charmrun)
Charm++> Running in Multicore mode: 12 threads (PEs)
Charm++> Using recursive bisection (scheme 3) for topology aware partitions
Converse/Charm++ Commit ID: v6.9.0-rc2-0-ge8d6e0c-namd-charm-6.9.0-build-2018-Oct-25-11430
Warning> Randomization of virtual memory (ASLR) is turned on in the kernel, thread migration may not work! Run 'echo 0 > /proc/sys/kernel/randomize_va_space' as root to disable it, or try running with '+isomalloc_sync'.
CharmLB> Load balancer assumes all CPUs are same.
Charm++> Running on 1 hosts (1 sockets x 12 cores x 1 PUs = 24-way SMP)
Charm++> Warning: Internally-determined PU count does not match hwloc's result!

History

#1 Updated by Sam White 19 days ago

  • Assignee set to Evan Ramos

Also available in: Atom PDF