update image repository and storage

kubernetes/helm/templates/ollama-statefulset.yaml

@@ -17,7 +17,10 @@ spec:
     spec:
       containers:
       - name: {{ include "ollama.name" . }}
-        image: {{ .Values.ollama.image }}
+        {{- with .Values.ollama.image }}
+        image: {{ .repository }}:{{ .tag }}
+        imagePullPolicy: {{ .pullPolicy }}
+        {{- end }}
         ports:
         - name: http
           containerPort: {{ .Values.ollama.service.containerPort }}
@@ -29,7 +32,7 @@ spec:
             value: /usr/local/nvidia/lib:/usr/local/nvidia/lib64
           - name: NVIDIA_DRIVER_CAPABILITIES
             value: compute,utility
-        {{- end}}
+        {{- end }}
         {{- with .Values.ollama.resources }}
         resources: {{- toYaml . | nindent 10 }}
         {{- end }}
@@ -41,11 +44,9 @@ spec:
       nodeSelector:
         {{- toYaml . | nindent 8 }}
       {{- end }}
+      {{- with .Values.ollama.tolerations }}
       tolerations:
-        {{- if .Values.ollama.gpu.enabled }}
+        {{- toYaml . | nindent 8 }}
-        - key: nvidia.com/gpu
-          operator: Exists
-          effect: NoSchedule
       {{- end }}
   volumeClaimTemplates:
   - metadata:
@@ -54,4 +55,4 @@ spec:
       accessModes: [ "ReadWriteOnce" ]
       resources:
         requests:
-          storage: {{ .Values.ollama.volumeSize }}
+          storage: {{ .Values.ollama.persistence.size }}

kubernetes/helm/templates/webui-deployment.yaml

@@ -16,7 +16,10 @@ spec:
     spec:
       containers:
       - name: {{ .Chart.Name }}
-        image: {{ .Values.webui.image }}
+        {{- with .Values.webui.image }}
+        image: {{ .repository }}:{{ .tag }}
+        imagePullPolicy: {{ .pullPolicy }}
+        {{- end }}
         ports:
         - name: http
           containerPort: {{ .Values.webui.service.containerPort }}

kubernetes/helm/values.yaml

@@ -1,6 +1,9 @@
 ollama:
   replicaCount: 1
-  image: ollama/ollama:latest
+  image:
+    repository: ollama/ollama
+    tag: latest
+    pullPolicy: Always
   resources:
     requests:
       cpu: "2000m"
@@ -9,9 +12,14 @@ ollama:
       cpu: "4000m"
       memory: "4Gi"
       nvidia.com/gpu: "0"
-  volumeSize: 30Gi
+  persistence:
+    enabled: true
+    size: 30Gi
   nodeSelector: {}
-  tolerations: []
+  tolerations:
+  - key: nvidia.com/gpu
+    operator: Exists
+    effect: NoSchedule
   service:
     type: ClusterIP
     port: 80
@@ -21,7 +29,10 @@ ollama:
 
 webui:
   replicaCount: 1
-  image: ghcr.io/open-webui/open-webui:main
+  image:
+    repository: ghcr.io/open-webui/open-webui
+    tag: main
+    pullPolicy: Always
   resources:
     requests:
       cpu: "500m"