limit ollama context

apps/ollama/release.yaml

@@ -19,13 +19,13 @@ spec:
       gpu:
         enabled: true
         type: nvidia
-        nvidiaResource: nvidia.com/gpu
         number: 1
       models:
         create:
           - name: qwen3.5:9b-custom
             template: |
               FROM qwen3.5:9b
+              PARAMETER num_ctx 8192
               PARAMETER temperature 1.0
               PARAMETER top_p 0.95
               PARAMETER top_k 20

infrastructure/controllers/nvidia-device-plugin/release.yaml

@@ -1,27 +0,0 @@
-apiVersion: helm.toolkit.fluxcd.io/v2
-kind: HelmRelease
-metadata:
-  name: nvidia-device-plugin
-  namespace: nvidia-device-plugin
-spec:
-  chart:
-    spec:
-      chart: nvidia-device-plugin
-      version: 0.17.x
-      sourceRef:
-        kind: HelmRepository
-        name: nvdp
-  interval: 15m
-  releaseName: nvidia-device-plugin
-  values:
-    runtimeClassName: nvidia
-    config:
-      default: time-slicing
-      map:
-        time-slicing: |
-          version: v1
-          sharing:
-            timeSlicing:
-              resources:
-              - name: nvidia.com/gpu
-                replicas: 5

infrastructure/controllers/nvidia-device-plugin/repository.yaml

@@ -1,8 +0,0 @@
-apiVersion: source.toolkit.fluxcd.io/v1
-kind: HelmRepository
-metadata:
-  name: nvdp
-  namespace: nvidia-device-plugin
-spec:
-  interval: 15m
-  url: https://nvidia.github.io/k8s-device-plugin

infrastructure/controllers/nvidia-device-plugin/runtime-class.yaml

@@ -1,5 +0,0 @@
-apiVersion: node.k8s.io/v1
-kind: RuntimeClass
-metadata:
-  name: nvidia
-handler: nvidia

infrastructure/controllers/nvidia-gpu-operator/release.yaml

@@ -0,0 +1,37 @@
+apiVersion: helm.toolkit.fluxcd.io/v2
+kind: HelmRelease
+metadata:
+  name: nvidia-gpu-operator
+  namespace: nvidia-gpu-operator
+spec:
+  chart:
+    spec:
+      chart: gpu-operator
+      version: 26.x
+      sourceRef:
+        kind: HelmRepository
+        name: nvidia
+  interval: 15m
+  releaseName: nvidia-gpu-operator
+  values:
+    driver:
+      enabled: false
+    toolkit:
+      enabled: false
+    hostPaths:
+      driverInstallDir: "/usr/local"
+    devicePlugin:
+      config:
+        create: true
+        name: "time-slicing-config"
+        default: "timeslicing"
+        data:
+          timeslicing: |-
+            version: v1
+            flags:
+              migStrategy: none
+            sharing:
+              timeSlicing:
+                resources:
+                - name: nvidia.com/gpu
+                  replicas: 5

infrastructure/controllers/nvidia-gpu-operator/repository.yaml

@@ -0,0 +1,8 @@
+apiVersion: source.toolkit.fluxcd.io/v1
+kind: HelmRepository
+metadata:
+  name: nvidia
+  namespace: nvidia-gpu-operator
+spec:
+  interval: 15m
+  url: https://helm.ngc.nvidia.com/nvidia

infrastructure/namespaces/namespace-nvidia-device-plugin.yaml

@@ -1,8 +0,0 @@
-apiVersion: v1
-kind: Namespace
-metadata:
-  name: nvidia-device-plugin
-  labels:
-    pod-security.kubernetes.io/enforce: privileged
-    pod-security.kubernetes.io/audit: privileged
-    pod-security.kubernetes.io/warn: privileged

infrastructure/namespaces/namespace-nvidia-gpu-operator.yaml

@@ -0,0 +1,6 @@
+apiVersion: v1
+kind: Namespace
+metadata:
+  name: nvidia-gpu-operator
+  labels:
+    pod-security.kubernetes.io/enforce: privileged