chore(test): update API version for nixl test

test/sidecar/README.md

@@ -8,7 +8,7 @@ a simple 1P1D sample application using the NIXL connector.
 To deploy this application in the `eval` cluster, run this command:
 
 ```
-$ kustomize build test/config/overlays/fmass/nixl | oc apply -f -
+$ kustomize build test/sidecar/config/overlays/llmd/nixl | oc apply -f -
 ```
 
 Wait a bit (up to 10mn) for the pods to be running.

test/sidecar/config/nixl/inferencepool.yaml

@@ -1,11 +1,17 @@
-apiVersion: inference.networking.x-k8s.io/v1alpha2
+apiVersion: inference.networking.k8s.io/v1
 kind: InferencePool
 metadata:
   name: qwen2-0-5b
 spec:
-  extensionRef:
-    name: qwen2-0-5b-epp
   selector:
-    llm-d.ai/inferenceServing: "true"
-    llm-d.ai/model: qwen2-0-5b
-  targetPortNumber: 8000
+    matchLabels:
+      llm-d.ai/inferenceServing: "true"
+      llm-d.ai/model: qwen2-0-5b
+  endpointPickerRef:
+    name: qwen2-0-5b-epp
+    kind: Service
+    port:
+      number: 9002
+  targetPorts:
+  - number: 8000
+

test/sidecar/config/nixl/kustomization.yaml

@@ -1,7 +1,6 @@
 resources:
   - qwen-decoder-pod.yaml
   - qwen-prefiller-pod.yaml
-  - inferencemodel.yaml
   - inferencepool.yaml
   - qwen-epp.yaml
   - qwen-epp-svc.yaml