Ticket 13525

Summary: In version 21.08, the slurmd -C command does not display the CPU topology according to whether Ignore_NUMA is configured
Product: Slurm Reporter: QinShuaiBing <qinshb>
Component: ConfigurationAssignee: Jacob Jenson <jacob>
Status: RESOLVED INVALID QA Contact:
Severity: 6 - No support contract    
Priority: ---    
Version: 21.08.0   
Hardware: Linux   
OS: Linux   
Site: -Other- Slinky Site: ---
Alineos Sites: --- Atos/Eviden Sites: ---
Confidential Site: --- Coreweave sites: ---
Cray Sites: --- DS9 clusters: ---
Google sites: --- HPCnow Sites: ---
HPE Sites: --- IBM Sites: ---
NOAA SIte: --- NoveTech Sites: ---
Nvidia HWinf-CS Sites: --- OCF Sites: ---
Recursion Pharma Sites: --- SFW Sites: ---
SNIC sites: --- Tzag Elita Sites: ---
Linux Distro: --- Machine Name:
CLE Version: Version Fixed:
Target Release: --- DevPrio: ---
Emory-Cloud Sites: ---

Description QinShuaiBing 2022-02-25 18:56:26 MST 
In version 21.08, when the slurmd -C command is executed, line 325 in the source code xcpuinfo.c, if (xstrcasestr(slurm_conf.sched_params, "Ignore_NUMA")), the structure slurm_conf.sched_params is not assigned at this time, so whether to configure Ignore_NUMA will not affect the results of slurmd -C, and has always obtained and displayed the CPU topology according to "Considering each L3Cache as a socket".
Is this a bug? Or did it on purpose?
Comment 1 Jacob Jenson 2022-02-28 09:41:41 MST 
Sugon will need to purchase a support contract before SchedMD can help you determine if this is a bug or on purpose.

Please contact sales@schedmd.com to start the process of purchasing support. 


Thank you, 
Jacob