Ticket 9980

Summary:	Cannot submit jobs to reservation if there is maint reservation
Product:	Slurm	Reporter:	CSC sysadmins <csc-slurm-tickets>
Component:	reservations	Assignee:	Dominik Bartkiewicz <bart>
Status:	RESOLVED FIXED	QA Contact:
Severity:	4 - Minor Issue
Priority:	---	CC:	alex
Version:	20.02.5
Hardware:	Linux
OS:	Linux
Site:	CSC - IT Center for Science	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:	20.02.6	Target Release:	---
DevPrio:	---	Emory-Cloud Sites:	---

Description CSC sysadmins 2020-10-13 05:47:17 MDT

Hi,

I was updating our compute node image and had a short maintenance window for node reboot. But users cannot submit jobs to the normal reservation and error message is quite hard to understand (at least user point of view). I think job should go to queue and wait if it cannot finish before the maint-reservation starts.


# scontrol create reservation=test starttime=now duration=14-00:00 users=tervotom nodes=c1170
# scontrol create reservation=test_maint StartTime=2020-10-13T15:00:00 duration=1:00:00 users=tervotom nodes=c1170 flags=maint


$ sbatch -A project_2001659 --reservation=test --nodes=1 -p medium -t 1:00:00 gpcnet_opmi_load.sh 
sbatch: error: Batch job submission failed: Requested node configuration is not available

BR,
Tommi

Comment 5 Dominik Bartkiewicz 2020-10-22 09:47:28 MDT

Hi

We've fixed this in commit
https://github.com/SchedMD/slurm/commit/67116e73 which will be in 20.02.6.

I'm closing this as resolved/fixed. Let us know if you have any more issues.

Dominik