feat: add nodesets specific authcred config

Co-authored-by: Fagani Hajizada <[email protected]>

Author: SkylerMalinowski

CHANGELOG.md

@@ -6,6 +6,8 @@ All notable changes to this project will be documented in this file.
 
 ### Added
 
+- Added NodeSet level authcred configuration.
+
 ### Fixed
 
 - Fixed token job `ttlSecondsAfterFinished` being too low for helm

helm/slurm/README.md

@@ -37,9 +37,10 @@ Helm Chart for Slurm HPC Workload Manager
 | compute.image.repository | string | `"ghcr.io/slinkyproject/slurmd"` |  Set the image repository to use. |
 | compute.image.tag | string | `"25.05-ubuntu24.04"` |  Set the image tag to use. |
 | compute.imagePullPolicy | string | `"IfNotPresent"` |  Set the image pull policy. |
-| compute.nodesets | list | `[{"affinity":{},"enabled":true,"extraVolumeMounts":[],"extraVolumes":[],"image":{"repository":"","tag":""},"imagePullPolicy":"IfNotPresent","name":"debug","nodeConfig":{},"nodeSelector":{"kubernetes.io/os":"linux"},"partition":{"config":{"MaxTime":"UNLIMITED","State":"UP"},"enabled":true},"persistentVolumeClaimRetentionPolicy":{"whenDeleted":"Retain","whenScaled":"Retain"},"priorityClassName":"","replicas":1,"resources":{},"tolerations":[],"updateStrategy":{"rollingUpdate":{"maxUnavailable":"20%"},"type":"RollingUpdate"},"useResourceLimits":true,"volumeClaimTemplates":[]}]` |  Slurm NodeSets by object list. |
-| compute.nodesets[0] | string | `{"affinity":{},"enabled":true,"extraVolumeMounts":[],"extraVolumes":[],"image":{"repository":"","tag":""},"imagePullPolicy":"IfNotPresent","name":"debug","nodeConfig":{},"nodeSelector":{"kubernetes.io/os":"linux"},"partition":{"config":{"MaxTime":"UNLIMITED","State":"UP"},"enabled":true},"persistentVolumeClaimRetentionPolicy":{"whenDeleted":"Retain","whenScaled":"Retain"},"priorityClassName":"","replicas":1,"resources":{},"tolerations":[],"updateStrategy":{"rollingUpdate":{"maxUnavailable":"20%"},"type":"RollingUpdate"},"useResourceLimits":true,"volumeClaimTemplates":[]}` |  Name of NodeSet. Must be unique. |
+| compute.nodesets | list | `[{"affinity":{},"authcred":{},"enabled":true,"extraVolumeMounts":[],"extraVolumes":[],"image":{"repository":"","tag":""},"imagePullPolicy":"IfNotPresent","name":"debug","nodeConfig":{},"nodeSelector":{"kubernetes.io/os":"linux"},"partition":{"config":{"MaxTime":"UNLIMITED","State":"UP"},"enabled":true},"persistentVolumeClaimRetentionPolicy":{"whenDeleted":"Retain","whenScaled":"Retain"},"priorityClassName":"","replicas":1,"resources":{},"tolerations":[],"updateStrategy":{"rollingUpdate":{"maxUnavailable":"20%"},"type":"RollingUpdate"},"useResourceLimits":true,"volumeClaimTemplates":[]}]` |  Slurm NodeSets by object list. |
+| compute.nodesets[0] | string | `{"affinity":{},"authcred":{},"enabled":true,"extraVolumeMounts":[],"extraVolumes":[],"image":{"repository":"","tag":""},"imagePullPolicy":"IfNotPresent","name":"debug","nodeConfig":{},"nodeSelector":{"kubernetes.io/os":"linux"},"partition":{"config":{"MaxTime":"UNLIMITED","State":"UP"},"enabled":true},"persistentVolumeClaimRetentionPolicy":{"whenDeleted":"Retain","whenScaled":"Retain"},"priorityClassName":"","replicas":1,"resources":{},"tolerations":[],"updateStrategy":{"rollingUpdate":{"maxUnavailable":"20%"},"type":"RollingUpdate"},"useResourceLimits":true,"volumeClaimTemplates":[]}` |  Name of NodeSet. Must be unique. |
 | compute.nodesets[0].affinity | object | `{}` |  Set affinity for Kubernetes Pod scheduling. Ref: https://kubernetes.io/docs/concepts/scheduling-eviction/assign-pod-node/#affinity-and-anti-affinity |
+| compute.nodesets[0].authcred | object | `{}` |  NodeSet specific authcred configuration, otherwise global authcred is used. |
 | compute.nodesets[0].enabled | bool | `true` |  Enables the NodeSet in Slurm. |
 | compute.nodesets[0].extraVolumeMounts | list | `[]` |  List of volume mounts. Ref: https://kubernetes.io/docs/concepts/storage/volumes/ |
 | compute.nodesets[0].extraVolumes | list | `[]` |  Define list of pod volumes. Ref: https://kubernetes.io/docs/concepts/storage/volumes/ |

helm/slurm/templates/compute/compute-nodeset.yaml

@@ -9,6 +9,19 @@ SPDX-License-Identifier: Apache-2.0
 {{- $tag := $nodeset.image.tag | default (include "slurm.compute.image.tag" $) }}
 {{- $imageRef := printf "%s:%s" $repository $tag }}
 {{- $name := printf "%s-%s" (include "slurm.compute.name" $) $nodeset.name }}
+{{- $authcredRepository := include "slurm.authcred.image.repository" $ }}
+{{- $authcredTag := include "slurm.authcred.image.repository" $ }}
+{{- $authcredImagePullPolicy := $nodeset.imagePullPolicy | default (include "slurm.imagePullPolicy" $) }}
+{{- $authcredResources := $.Values.authcred.resources }}
+{{- with $nodeset.authcred -}}
+  {{- with .image -}}
+    {{- $authcredRepository = .repository | default $authcredRepository -}}
+    {{- $authcredTag = .tag | default $authcredTag -}}
+  {{- end -}}
+  {{- $authcredImagePullPolicy = .imagePullPolicy | default $authcredImagePullPolicy -}}
+  {{- $authcredResources = .resources | default $authcredResources -}}
+{{- end -}}
+{{- $authcredImageRef := printf "%s:%s" $authcredRepository $authcredTag }}
 ---
 apiVersion: slinky.slurm.net/v1alpha1
 kind: NodeSet
@@ -50,12 +63,12 @@ spec:
       {{- include "slurm.imagePullSecrets" $ | nindent 6 }}
       initContainers:
         - name: init
-          image: {{ include "slurm.authcred.imageRef" $ }}
-          imagePullPolicy: {{ $.Values.authcred.imagePullPolicy | default (include "slurm.imagePullPolicy" $) }}
-          {{- with $.Values.authcred.resources }}
+          image: {{ $authcredImageRef }}
+          imagePullPolicy: {{ $authcredImagePullPolicy }}
+          {{- with $authcredResources }}
           resources:
             {{- toYaml . | nindent 12 }}
-          {{- end }}{{- /* with .Values.authcred.resources */}}
+          {{- end }}{{- /* with $authcredResources */}}
           env:
             - name: SLURM_USER
               value: {{ include "slurm.user" $ }}
@@ -72,8 +85,8 @@ spec:
           volumeMounts:
             {{- include "slurm.init.volumeMounts" $ | nindent 12 }}
         - name: logfile
-          image: {{ include "slurm.authcred.imageRef" $ }}
-          imagePullPolicy: {{ $.Values.authcred.imagePullPolicy | default (include "slurm.imagePullPolicy" $) }}
+          image: {{ $authcredImageRef }}
+          imagePullPolicy: {{ $authcredImagePullPolicy }}
           restartPolicy: Always
           env:
             - name: SOCKET
@@ -88,10 +101,10 @@ spec:
               {{- range $.Files.Lines "scripts/socket-logger.sh" }}
               {{ . }}
               {{- end }}{{- /* range $.Files.Lines "scripts/socket-logger.sh" */}}
-          {{- with $.Values.authcred.resources }}
+          {{- with $authcredResources }}
           resources:
             {{- toYaml . | nindent 12 }}
-          {{- end }}{{- /* with $.Values.authcred.resources */}}
+          {{- end }}{{- /* with $authcredResources */}}
           volumeMounts:
             - name: slurm-log
               mountPath: /var/log/slurm

helm/slurm/values.yaml

@@ -512,6 +512,15 @@ compute:
         # Set the image tag to use.
         tag: ""
       #
+      # -- (object)
+      # NodeSet specific authcred configuration, otherwise global authcred is used.
+      authcred: {}
+        # imagePullPolicy: IfNotPresent
+        # image:
+        #   repository: ghcr.io/slinkyproject/sackd
+        #   tag: 25.05-ubuntu24.04
+        # resources: {}
+      #
       # -- (string)
       # Set the priority class to use.
       # Ref: https://kubernetes.io/docs/concepts/scheduling-eviction/pod-priority-preemption/#priorityclass