Ticket 2635

Summary:	15.08.10 incompatible with bf_min_prio_reserve patch
Product:	Slurm	Reporter:	Doug Jacobsen <dmjacobsen>
Component:	slurmctld	Assignee:	Director of Support <support>
Status:	RESOLVED DUPLICATE	QA Contact:
Severity:	3 - Medium Impact
Priority:	---
Version:	15.08.10
Hardware:	Cray XC
OS:	Linux
Site:	NERSC	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 Doug Jacobsen 2016-04-13 07:14:45 MDT

Hello,

I just tried to upgrade slurm on cori to 15.08.10 but it crashed as soon as I brought up the "regular" partition owing to some incompatibilities between the bf_min_prio_reserve patch we have and some modifications made in 15.08.10 reverting nearby logic.


In particular the segfault occurred in backfill.c:

                job_ptr->part_ptr = part_ptr;
                job_ptr->priority = job_queue_rec->priority;

when job_queue_rec is accessed after it is xfreed() earlier.

I both really need the bf_min_prio_reserve and I need to be able to upgrade 15.08.  Would it be possible that we can get an update of this patch?  or better, get bf_min_prio_reserve mainlined into 15.08 HEAD?

I'm attaching the mildly hacked up patch I'm using on top of 15.08.10.

I'm in a bit of a jam on cori now, because we are up, but I can't redeploy or get some of the other patches unless I revert changes to my build system...


Thank you for your help with this,
Doug

Comment 1 Doug Jacobsen 2016-04-14 00:58:46 MDT

duplicate of 2634

*** This ticket has been marked as a duplicate of ticket 2634 ***