fix: Resolve quickstart script failures and add automated testing

.github/workflows/quickstart-integration-test.yml

@@ -0,0 +1,110 @@
+name: Quickstart Integration Test
+
+on:
+  pull_request:
+    branches:
+      - main
+    paths:
+      - 'scripts/quickstart.sh'
+      - 'deploy/docker-compose/**'
+      - 'config/config.yaml'
+      - 'tools/make/common.mk'
+      - 'tools/make/models.mk'
+      - 'tools/make/docker.mk'
+  workflow_dispatch: # Allow manual triggering
+
+jobs:
+  test-quickstart:
+    runs-on: ubuntu-latest
+    timeout-minutes: 30
+
+    steps:
+      - name: Check out the repo
+        uses: actions/checkout@v4
+
+      - name: Free up disk space
+        run: |
+          echo "Disk space before cleanup:"
+          df -h
+          sudo rm -rf /usr/share/dotnet
+          sudo rm -rf /opt/ghc
+          sudo rm -rf /usr/local/share/boost
+          sudo rm -rf "$AGENT_TOOLSDIRECTORY"
+          echo "Disk space after cleanup:"
+          df -h
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.11'
+
+      - name: Install system dependencies
+        run: |
+          sudo apt-get update
+          sudo apt-get install -y \
+            make \
+            curl \
+            docker-compose
+
+      - name: Run quickstart script
+        id: quickstart
+        run: |
+          timeout 1200 bash scripts/quickstart.sh || {
+            exit_code=$?
+            if [ $exit_code -eq 124 ]; then
+              echo "::error::Quickstart script timed out after 20 minutes"
+            else
+              echo "::error::Quickstart script failed with exit code $exit_code"
+            fi
+            exit $exit_code
+          }
+        env:
+          CI: true
+          CI_MINIMAL_MODELS: true
+          TERM: xterm
+          HF_HUB_ENABLE_HF_TRANSFER: 1
+          HF_HUB_DISABLE_TELEMETRY: 1
+
+      - name: Test semantic routing functionality
+        run: |
+          echo "Testing semantic router with a sample query..."
+
+          response=$(curl -s -X POST http://localhost:8801/v1/chat/completions \
+            -H "Content-Type: application/json" \
+            -d '{
+              "model": "qwen3",
+              "messages": [{"role": "user", "content": "What is 2 + 2?"}],
+              "temperature": 0.7
+            }')
+
+          echo "Full response: $response"
+
+          # Validate response structure
+          if echo "$response" | jq -e '.choices[0].message.content' > /dev/null 2>&1; then
+            echo "✓ Semantic router successfully routed and processed the query"
+            echo "  Answer: $(echo "$response" | jq -r '.choices[0].message.content' | head -c 200)"
+          else
+            echo "::error::Semantic router failed to process query correctly"
+            echo "Response was: $response"
+            exit 1
+          fi
+
+      - name: Show service logs on failure
+        if: failure()
+        run: |
+          echo "=== Docker Compose Logs ==="
+          docker compose -f deploy/docker-compose/docker-compose.yml logs
+          echo "=== Container Status ==="
+          docker ps -a
+          echo "=== Semantic Router Logs ==="
+          docker logs semantic-router || true
+          echo "=== Envoy Logs ==="
+          docker logs envoy-proxy || true
+          echo "=== Dashboard Logs ==="
+          docker logs semantic-router-dashboard || true
+
+      - name: Clean up
+        if: always()
+        run: |
+          make docker-compose-down || true
+          docker system prune -af --volumes || true

config/config.yaml

@@ -24,15 +24,6 @@ semantic_cache:
   # Options: "bert" (fast, 384-dim), "qwen3" (high quality, 1024-dim, 32K context), "gemma" (balanced, 768-dim, 8K context)
   # Default: "bert" (fastest, lowest memory)
   embedding_model: "bert"
-  # HNSW index configuration (for memory backend only)
-  use_hnsw: true # Enable HNSW index for faster similarity search
-  hnsw_m: 16 # Number of bi-directional links (higher = better recall, more memory)
-  hnsw_ef_construction: 200 # Construction parameter (higher = better quality, slower build)
-
-  # Hybrid cache configuration (when backend_type: "hybrid")
-  # Combines in-memory HNSW for fast search with Milvus for scalable storage
-  # max_memory_entries: 100000 # Max entries in HNSW index (default: 100,000)
-  # backend_config_path: "config/milvus.yaml" # Path to Milvus config
 
 tools:
   enabled: true
@@ -223,7 +214,7 @@ router:
   traditional_attention_dropout_prob: 0.1          # Traditional model attention dropout probability
   tie_break_confidence: 0.5                        # Confidence value for tie-breaking situations
 
-default_model: openai/gpt-oss-20b
+default_model: qwen3
 
 # Reasoning family configurations
 reasoning_families:

deploy/docker-compose/docker-compose.yml

@@ -7,9 +7,9 @@ services:
     ports:
       - "50051:50051"
     volumes:
-      - ../../config:/app/config:ro
-      - ../../models:/app/models:ro
-      - ~/.cache/huggingface:/root/.cache/huggingface
+      - ../../config:/app/config:ro,z
+      - ../../models:/app/models:ro,z
+      - ~/.cache/huggingface:/root/.cache/huggingface:z
     environment:
       - LD_LIBRARY_PATH=/app/lib
       # Use main config by default; override via CONFIG_FILE if needed
@@ -32,19 +32,21 @@ services:
   envoy:
     image: envoyproxy/envoy:v1.31.7
     container_name: envoy-proxy
+    security_opt:
+      - label=disable
     ports:
       - "8801:8801"  # Main proxy port
       - "19000:19000"  # Admin interface
     volumes:
-      - ./addons/envoy.yaml:/etc/envoy/envoy.yaml:ro
+      - ./addons/envoy.yaml:/etc/envoy/envoy.yaml:ro,z
     command: ["/usr/local/bin/envoy", "-c", "/etc/envoy/envoy.yaml", "--component-log-level", "ext_proc:trace,router:trace,http:trace"]
     depends_on:
       semantic-router:
         condition: service_healthy
     networks:
       - semantic-network
     healthcheck:
-      test: ["CMD", "curl", "-f", "http://localhost:19000/ready"]
+      test: ["CMD", "bash", "-c", "(echo -e 'GET /ready HTTP/1.1\\r\\nHost: localhost\\r\\n\\r\\n' >&3; timeout 2 cat <&3) 3<>/dev/tcp/localhost/19000 | grep -q LIVE"]
       interval: 10s
       timeout: 5s
       retries: 5
@@ -86,7 +88,7 @@ services:
     image: prom/prometheus:v2.53.0
     container_name: prometheus
     volumes:
-      - ./addons/prometheus.yaml:/etc/prometheus/prometheus.yaml:ro
+      - ./addons/prometheus.yaml:/etc/prometheus/prometheus.yaml:ro,z
       - prometheus-data:/prometheus
     command:
       - --config.file=/etc/prometheus/prometheus.yaml
@@ -106,11 +108,11 @@ services:
     ports:
       - "3000:3000"
     volumes:
-      - ./addons/grafana.ini:/etc/grafana/grafana.ini:ro
-      - ./addons/grafana-datasource.yaml:/etc/grafana/provisioning/datasources/datasource.yaml:ro
-      - ./addons/grafana-datasource-jaeger.yaml:/etc/grafana/provisioning/datasources/datasource_jaeger.yaml:ro
-      - ./addons/grafana-dashboard.yaml:/etc/grafana/provisioning/dashboards/dashboard.yaml:ro
-      - ./addons/llm-router-dashboard.json:/etc/grafana/provisioning/dashboards/llm-router-dashboard.json:ro
+      - ./addons/grafana.ini:/etc/grafana/grafana.ini:ro,z
+      - ./addons/grafana-datasource.yaml:/etc/grafana/provisioning/datasources/datasource.yaml:ro,z
+      - ./addons/grafana-datasource-jaeger.yaml:/etc/grafana/provisioning/datasources/datasource_jaeger.yaml:ro,z
+      - ./addons/grafana-dashboard.yaml:/etc/grafana/provisioning/dashboards/dashboard.yaml:ro,z
+      - ./addons/llm-router-dashboard.json:/etc/grafana/provisioning/dashboards/llm-router-dashboard.json:ro,z
       - grafana-data:/var/lib/grafana
     networks:
       - semantic-network
@@ -175,9 +177,16 @@ services:
       - PYTHONUNBUFFERED=1
     volumes:
       # Persistent pipelines storage (auto-loaded on start)
-      - openwebui-pipelines:/app/pipelines
-      # Mount our vLLM Semantic Router pipeline
-      - ./addons/vllm_semantic_router_pipe.py:/app/pipelines/vllm_semantic_router_pipe.py:ro
+      - type: volume
+        source: openwebui-pipelines
+        target: /app/pipelines
+        volume:
+          nocopy: true
+      # Mount our vLLM Semantic Router pipeline (read-only) into the persistent dir
+      - type: bind
+        source: ./addons/vllm_semantic_router_pipe.py
+        target: /app/pipelines/vllm_semantic_router_pipe.py
+        read_only: true
     networks:
       - semantic-network
 
@@ -202,7 +211,7 @@ services:
       - HUGGINGFACE_HUB_TOKEN=${HUGGINGFACE_HUB_TOKEN:-}
       - HF_HUB_ENABLE_HF_TRANSFER=1
     volumes:
-      - ../../models:/app/models:ro
+      - ../../models:/app/models:ro,z
       - hf-cache:/home/llmkatan/.cache/huggingface
     networks:
       semantic-network:
@@ -235,7 +244,7 @@ services:
       - TARGET_CHATUI_URL=http://chat-ui:3000
       - ROUTER_CONFIG_PATH=/app/config/config.yaml
     volumes:
-      - ../../config:/app/config:rw
+      - ../../config:/app/config:rw,z
     ports:
       - "8700:8700"
     networks: