feat: vLLM Prometheus metrics scraping (#37)

Add ServiceMonitor for llm-serving namespace. Wire prometheus.io annotations to all vLLM pods (reasoning, ornith, embeddings, reranker). Scrape /metrics@8080 every 30s with proper relabeling.

k8s/apps/llm-serving/embeddings.yaml

@@ -45,6 +45,14 @@ spec:
       volumeMounts:
       - mountPath: /mnt/models
         name: models
+    podMetadata:
+      annotations:
+        prometheus.io/scrape: "true"
+        prometheus.io/port: "8080"
+        prometheus.io/path: "/metrics"
+      labels:
+        app.kubernetes.io/name: llm-embeddings
+        app.kubernetes.io/part-of: llm-serving
     maxReplicas: 1
     minReplicas: 1
     nodeSelector:

k8s/apps/llm-serving/ornith.yaml

@@ -83,6 +83,14 @@ spec:
       volumeMounts:
       - mountPath: /mnt/models
         name: models
+    podMetadata:
+      annotations:
+        prometheus.io/scrape: "true"
+        prometheus.io/port: "8080"
+        prometheus.io/path: "/metrics"
+      labels:
+        app.kubernetes.io/name: llm-ornith
+        app.kubernetes.io/part-of: llm-serving
     deploymentStrategy:
       type: Recreate
     # 1 replica -- ornith:35b only. qwen2.5:3b moved to CPU on cp-2.

k8s/apps/llm-serving/reasoning.yaml

@@ -104,6 +104,14 @@ spec:
         name: models
       - mountPath: /dev/shm
         name: shm
+    podMetadata:
+      annotations:
+        prometheus.io/scrape: "true"
+        prometheus.io/port: "8080"
+        prometheus.io/path: "/metrics"
+      labels:
+        app.kubernetes.io/name: llm-reasoning
+        app.kubernetes.io/part-of: llm-serving
     deploymentStrategy:
       type: Recreate
     maxReplicas: 1

k8s/apps/llm-serving/reranker.yaml

@@ -45,6 +45,14 @@ spec:
       volumeMounts:
       - mountPath: /mnt/models
         name: models
+    podMetadata:
+      annotations:
+        prometheus.io/scrape: "true"
+        prometheus.io/port: "8080"
+        prometheus.io/path: "/metrics"
+      labels:
+        app.kubernetes.io/name: llm-reranker
+        app.kubernetes.io/part-of: llm-serving
     maxReplicas: 1
     minReplicas: 1
     nodeSelector: