3.13.2 config file error upon startup for UnkillableStepTimeout and MessageTimeout

[gwolski]

Using Rocky8.10/APC 3.13.2.

On head node and compute node start up there are unsettling error messages:

headnode:
/var/log/messages-20250713:Jul 8 18:04:35 ip-10-6-6-5 slurmctld[1199686]: slurmctld: error: UnkillableStepTimeout must be at least 5 times greater than MessageTimeout, otherwise nodes may go down with the reason "KillTaskFailed". Current values: UnkillableStepTimeout=180, MessageTimeout=60

Compute node:
Jul 19 16:06:24 ip-10-6-4-238 slurmd[4569]: slurmd: error: UnkillableStepTimeout must be at least 5 times greater than MessageTimeout, otherwise nodes may go down with the reason "KillTaskFailed". Current values: UnkillableStepTimeout=180, MessageTimeout=60
Jul 19 16:06:24 ip-10-6-4-238 slurmd[4569]: error: UnkillableStepTimeout must be at least 5 times greater than MessageTimeout, otherwise nodes may go down with the reason "KillTaskFailed". Current values: UnkillableStepTimeout=180, MessageTimeout=60

[hanwen-cluster]

Hi @gwolski

We have merged a fix PR aws/aws-parallelcluster-cookbook#2913.

The fix will be in next release.

So far, we are not aware of any problems caused by this parameter.

Thank you