Update

tasks/main.yml

@@ -21,6 +21,7 @@
       deployment:
         image:
           tag: latest-gpu-nvidia-cuda-12
+        pullPolicy: Always
         runtimeClassName: nvidia
         secretEnv:
           - name: LOCALAI_DISABLE_WEBUI
@@ -33,6 +34,12 @@
             value: "{{ localai_watchdog_idle_timeout }}"
           - name: LOCALAI_WATCHDOG_BUSY
             value: "{{ localai_watchdog_busy }}"
+          - name: DEBUG
+            value: "{{ localai_debug }}"
+          - name: LOCALAI_P2P
+            value: "{{ localai_p2p }}"
+          - name: LOCALAI_FEDERATED
+            value: "{{ localai_federated }}"
       resources:
         limits:
           nvidia.com/gpu: 1
@@ -46,14 +53,13 @@
           operator: Equal
           value: GPU
           effect: NoSchedule
-      service:
-        externalTrafficPolicy: Local
       ingress:
         enabled: true
         className: nginx
         annotations:
           cert-manager.io/cluster-issuer: ca-issuer
           nginx.ingress.kubernetes.io/proxy-read-timeout: "600"
+          nginx.ingress.kubernetes.io/proxy-set-header: "X-Forwarded-For"
         hosts:
           - host: localai.eom.dev
             paths: