From 8b4536b41d18585dd5ae96e389b285d25d4dc2a6 Mon Sep 17 00:00:00 2001
From: Stephan <31624652+stephandooper@users.noreply.github.com>
Date: Wed, 9 Aug 2023 10:10:55 +0200
Subject: [PATCH] Update slurm.conf

Uncommented AccountingStorageEnforce
Added UnkillableStepTimeout to accomodate older water-cooled machines that take a longer time to start/kill jobs
Adjusted killwait from 30 to 90 for the same reason.
---
 roles/slurm/templates/etc/slurm/slurm.conf | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/roles/slurm/templates/etc/slurm/slurm.conf b/roles/slurm/templates/etc/slurm/slurm.conf
index 70eb837ac..3f4de7f2a 100644
--- a/roles/slurm/templates/etc/slurm/slurm.conf
+++ b/roles/slurm/templates/etc/slurm/slurm.conf
@@ -88,8 +88,9 @@ SlurmctldTimeout=120
 SlurmdTimeout=300
 InactiveLimit=0
 MinJobAge=300
-KillWait=30
+KillWait=90
 Waittime=0
+UnkillableStepTimeout=180
 
 # SCHEDULING
 SchedulerType=sched/backfill
@@ -123,7 +124,7 @@ AccountingStorageTRES=gres/gpu
 AccountingStorageType=accounting_storage/slurmdbd
 AccountingStorageHost={{ groups["slurm-master"][0] }}
 #AccountingStorageLoc=
-#AccountingStorageEnforce=associations,limits,qos
+AccountingStorageEnforce=associations,limits,qos
 AccountingStorageUser={{ slurm_db_username }}
 AccountingStoragePass=/var/run/munge/munge.socket.2