expand pvc size & fix inference-pool selector error

Signed-off-by: JaredforReal <w13431838023@gmail.com>

Author: JaredforReal

deploy/kubernetes/README.md

@@ -7,7 +7,7 @@ This directory contains Kubernetes manifests for deploying the Semantic Router u
 The deployment consists of:
 
 - **ConfigMap**: Contains `config.yaml` and `tools_db.json` configuration files
-- **PersistentVolumeClaim**: 10Gi storage for model files
+- **PersistentVolumeClaim**: 30Gi storage for model files (adjust based on models you enable)
 - **Deployment**:
   - **Init Container**: Downloads/copies model files to persistent volume
   - **Main Container**: Runs the semantic router service
@@ -29,11 +29,11 @@ The deployment consists of:
 kubectl apply -k deploy/kubernetes/
 
 # Check deployment status
-kubectl get pods -l app=semantic-router -n semantic-router
-kubectl get services -l app=semantic-router -n semantic-router
+kubectl get pods -l app=semantic-router -n vllm-semantic-router-system
+kubectl get services -l app=semantic-router -n vllm-semantic-router-system
 
 # View logs
-kubectl logs -l app=semantic-router -n semantic-router -f
+kubectl logs -l app=semantic-router -n vllm-semantic-router-system -f
 ```
 
 ### Kind (Kubernetes in Docker) Deployment
@@ -86,20 +86,20 @@ kubectl wait --for=condition=Ready nodes --all --timeout=300s
 kubectl apply -k deploy/kubernetes/
 
 # Wait for deployment to be ready
-kubectl wait --for=condition=Available deployment/semantic-router -n semantic-router --timeout=600s
+kubectl wait --for=condition=Available deployment/semantic-router -n vllm-semantic-router-system --timeout=600s
 ```
 
 **Step 3: Check deployment status**
 
 ```bash
 # Check pods
-kubectl get pods -n semantic-router -o wide
+kubectl get pods -n vllm-semantic-router-system -o wide
 
 # Check services
-kubectl get services -n semantic-router
+kubectl get services -n vllm-semantic-router-system
 
 # View logs
-kubectl logs -l app=semantic-router -n semantic-router -f
+kubectl logs -l app=semantic-router -n vllm-semantic-router-system -f
 ```
 
 #### Resource Requirements for Kind
@@ -137,13 +137,13 @@ Or using kubectl directly:
 
 ```bash
 # Access Classification API (HTTP REST)
-kubectl port-forward -n semantic-router svc/semantic-router 8080:8080
+kubectl port-forward -n vllm-semantic-router-system svc/semantic-router 8080:8080
 
 # Access gRPC API
-kubectl port-forward -n semantic-router svc/semantic-router 50051:50051
+kubectl port-forward -n vllm-semantic-router-system svc/semantic-router 50051:50051
 
 # Access metrics
-kubectl port-forward -n semantic-router svc/semantic-router-metrics 9190:9190
+kubectl port-forward -n vllm-semantic-router-system svc/semantic-router-metrics 9190:9190
 ```
 
 #### Testing the Deployment
@@ -195,6 +195,11 @@ kubectl delete -k deploy/kubernetes/
 kind delete cluster --name semantic-router-cluster
 ```
 
+## Notes on dependencies
+
+- Gateway API Inference Extension CRDs are required only when using the Envoy AI Gateway integration in `deploy/kubernetes/ai-gateway/`. Follow the installation steps in `website/docs/installation/kubernetes.md` if you plan to use the gateway path.
+- The core kustomize deployment in this folder does not install Envoy Gateway or AI Gateway; those are optional components documented separately.
+
 ## Make Commands Reference
 
 The project provides comprehensive make targets for managing kind clusters and deployments:
@@ -293,6 +298,11 @@ kubectl top pods -n semantic-router
 # Adjust resource limits in deployment.yaml if needed
 ```
 
+### Storage sizing
+
+- The default PVC is 30Gi. If the enabled models are small, you can reduce it; otherwise reserve at least 2–3x the total model size.
+- If your cluster's default StorageClass isn't named `standard`, change `storageClassName` in `pvc.yaml` accordingly or remove the field to use the default class.
+
 ### Resource Optimization
 
 For different environments, you can adjust resource requirements:

deploy/kubernetes/ai-gateway/inference-pool/inference-pool.yaml

@@ -11,7 +11,7 @@ spec:
     - number: 50051
   selector:
     matchLabels:
-      app: vllm-semantic-router
+      app: semantic-router
   endpointPickerRef:
     name: semantic-router
     port:

deploy/kubernetes/pvc.yaml

@@ -9,5 +9,5 @@ spec:
     - ReadWriteOnce
   resources:
     requests:
-      storage: 10Gi
+      storage: 30Gi
   storageClassName: standard

website/docs/installation/kubernetes.md

@@ -37,6 +37,13 @@ kubectl wait --for=condition=Ready nodes --all --timeout=300s
 
 Configure the semantic router by editing `deploy/kubernetes/config.yaml`. This file contains the vLLM configuration, including model config, endpoints, and policies.
 
+Important notes before you apply manifests:
+
+- `vllm_endpoints.address` must be an IP address (not hostname) reachable from inside the cluster. If your LLM backends run as K8s Services, use the ClusterIP (for example `10.96.0.10`) and set `port` accordingly. Do not include protocol or path.
+- The PVC in `deploy/kubernetes/pvc.yaml` uses `storageClassName: standard`. On some clouds or local clusters, the default StorageClass name may differ (e.g., `standard-rwo`, `gp2`, or a provisioner like local-path). Adjust as needed.
+- Default PVC size is 30Gi. Size it to at least 2–3x of your total model footprint to leave room for indexes and updates.
+- The initContainer downloads several models from Hugging Face on first run and writes them into the PVC. Ensure outbound egress to Hugging Face is allowed and there is at least ~6–8 GiB free space for the models specified.
+
 Deploy the semantic router service with all required components:
 
 ```bash
@@ -135,26 +142,28 @@ Expected output should show the inference pool in `Accepted` state:
 ```yaml
 status:
   parent:
-  - conditions:
-    - lastTransitionTime: "2025-09-27T09:27:32Z"
-      message: 'InferencePool has been Accepted by controller ai-gateway-controller:
-        InferencePool reconciled successfully'
-      observedGeneration: 1
-      reason: Accepted
-      status: "True"
-      type: Accepted
-    - lastTransitionTime: "2025-09-27T09:27:32Z"
-      message: 'Reference resolution by controller ai-gateway-controller: All references
-        resolved successfully'
-      observedGeneration: 1
-      reason: ResolvedRefs
-      status: "True"
-      type: ResolvedRefs
-    parentRef:
-      group: gateway.networking.k8s.io
-      kind: Gateway
-      name: vllm-semantic-router
-      namespace: vllm-semantic-router-system
+    - conditions:
+        - lastTransitionTime: "2025-09-27T09:27:32Z"
+          message:
+            "InferencePool has been Accepted by controller ai-gateway-controller:
+            InferencePool reconciled successfully"
+          observedGeneration: 1
+          reason: Accepted
+          status: "True"
+          type: Accepted
+        - lastTransitionTime: "2025-09-27T09:27:32Z"
+          message:
+            "Reference resolution by controller ai-gateway-controller: All references
+            resolved successfully"
+          observedGeneration: 1
+          reason: ResolvedRefs
+          status: "True"
+          type: ResolvedRefs
+      parentRef:
+        group: gateway.networking.k8s.io
+        kind: Gateway
+        name: vllm-semantic-router
+        namespace: vllm-semantic-router-system
 ```
 
 ## Testing the Deployment