Ticket 7355

Summary:	excessive slurmctld starting time
Product:	Slurm	Reporter:	IDRIS System Team <gensyshpe>
Component:	slurmctld	Assignee:	Marcin Stolarek <cinek>
Status:	RESOLVED INVALID	QA Contact:
Severity:	3 - Medium Impact
Priority:	---	CC:	cinek, nate
Version:	18.08.7
Hardware:	Linux
OS:	Linux
Site:	IDRIS	Alineos Sites:	---
Atos/Eviden Sites:	---	Confidential Site:	---
Coreweave sites:	---	Cray Sites:	---
DS9 clusters:	---	HPCnow Sites:	---
HPE Sites:	---	IBM Sites:	---
NOAA SIte:	---	OCF Sites:	---
Recursion Pharma Sites:	---	SFW Sites:	---
SNIC sites:	---	Linux Distro:	---
Machine Name:		CLE Version:
Version Fixed:		Target Release:	---
DevPrio:	---	Emory-Cloud Sites:	---

Description IDRIS System Team 2019-07-04 04:08:40 MDT

Hi,
We are installing Slurm 18.08.7 on a large configuration (about 1800 compute nodes) and we noticed that slurmctld takes quite a long time (around 30 minutes) to start or to re-read the configuration file slurm.conf .
Is this the starting time we should expect or could it be the result of a problem?
What can we do to make slurmctld start faster ?

Regards,

Philipe  Collinet

Comment 1 Marcin Stolarek 2019-07-04 07:48:29 MDT

Could you please share slurmctld log from start time with us with debug2 enabled? 

cheers,
Marcin

Comment 4 IDRIS System Team 2019-07-08 05:24:13 MDT

Hello,

 We setup the debug mode and it helped US.
 We found that the adress resolution of the node running the slurmctld was node correct and had to wait for the second adress to succed the resolution.

  Whence this issue corrected, the starting time is correct.

  Thanks for your help. We can close the Bug.

Best regards,

Philippe Collinet

Comment 5 Marcin Stolarek 2019-07-08 05:35:30 MDT

Thanks for the update. I'm closing this bug report as invalid.

cheers,
Marcin