Merge pull request #9 from rwilliamspbg-ops/feat/model-library-worker-reconnect-compose

Feat/model library worker reconnect compose

Author: rwilliamspbg-ops

.github/workflows/ci.yml

@@ -86,3 +86,57 @@ jobs:
       - name: Build Docker images
         run: |
           docker compose -f docker-compose.dev.yml build
+
+  compose-smoke:
+    needs: build
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Start compose stack
+        run: |
+          docker compose -f docker-compose.yml up -d --build
+
+      - name: Wait for services
+        run: |
+          python - <<'PY'
+          import time
+          import urllib.request
+
+          urls = [
+              "http://127.0.0.1:8003/health",
+              "http://127.0.0.1:8004/health",
+              "http://127.0.0.1:8003/api/workers",
+              "http://127.0.0.1:8003/api/metrics",
+          ]
+
+          deadline = time.time() + 180
+          while time.time() < deadline:
+              try:
+                  for url in urls:
+                      with urllib.request.urlopen(url, timeout=5) as response:
+                          if response.status != 200:
+                              raise RuntimeError(f"{url} returned {response.status}")
+                  break
+              except Exception:
+                  time.sleep(5)
+          else:
+              raise SystemExit("compose services did not become healthy in time")
+          PY
+
+      - name: Verify GUI-facing endpoints
+        run: |
+          curl -fsS http://127.0.0.1:8003/health
+          curl -fsS http://127.0.0.1:8003/api/workers
+          curl -fsS http://127.0.0.1:8003/api/metrics
+          curl -fsS http://127.0.0.1:8004/health
+
+      - name: Dump compose logs on failure
+        if: failure()
+        run: |
+          docker compose -f docker-compose.yml logs --no-color
+
+      - name: Tear down compose stack
+        if: always()
+        run: |
+          docker compose -f docker-compose.yml down -v

Dockerfile

@@ -15,9 +15,11 @@ RUN apt-get update && apt-get install -y     gcc     git     libgl1     libegl1
 # Copy requirements first for better caching
 COPY requirements.txt .
 RUN pip install --no-cache-dir -r requirements.txt
+RUN pip install --no-cache-dir "fastapi>=0.104.0" "uvicorn>=0.24.0" "requests>=2.31.0"
 
 # Copy application code
 COPY mohawk_gui/ ./mohawk_gui/
+COPY prototype/ ./prototype/
 
 # Create non-root user for security
 RUN groupadd mohawk && useradd -r -g mohawk mohawk

Dockerfile.controller

@@ -37,11 +37,11 @@ RUN chown -R appuser:appuser /app
 USER appuser
 
 # Expose controller port
-EXPOSE 9000
+EXPOSE 8000
 
 # Health check
 HEALTHCHECK --interval=30s --timeout=10s --retries=3 \
-    CMD curl -f http://localhost:9000/health || exit 1
+    CMD curl -f http://localhost:8000/health || exit 1
 
-# Run controller (in production, would use proper entrypoint)
-CMD ["python", "-c", "from prototype.controller import Controller; print('Controller ready')"]
+# Run a lightweight controller API used for local compose verification.
+CMD ["python", "-m", "uvicorn", "prototype.controller_service:app", "--host", "0.0.0.0", "--port", "8000"]

docker-compose.dev.yml

@@ -7,7 +7,7 @@ services:
       - "8000:8000"
     environment:
       - WORKER_URLS=http://worker1:8000,http://worker2:8000
-    command: python -m http.server 8000
+    command: python -m uvicorn prototype.controller_service:app --host 0.0.0.0 --port 8000
     depends_on:
       - worker1
       - worker2

docker-compose.yml

@@ -30,8 +30,8 @@ services:
       - PYTHONUNBUFFERED=1
       - PYTHONDONTWRITEBYTECODE=1
       - QT_QPA_PLATFORM=offscreen
-    # Lightweight API backend used by the desktop GUI for local verification.
-    command: python -m uvicorn mohawk_gui.mock_backend:app --host 0.0.0.0 --port 8003
+    # Real controller API backend used by the desktop GUI.
+    command: python -m uvicorn prototype.controller_service:app --host 0.0.0.0 --port 8003
     healthcheck:
       test: ["CMD", "python", "-c", "import urllib.request; urllib.request.urlopen('http://127.0.0.1:8003/health', timeout=3)"]
       interval: 30s

mohawk/models/loader.py

@@ -87,7 +87,7 @@ def add_local_model(self, model_path: str, alias: Optional[str] = None) -> Dict[
         model_id = alias or path.name
         return self.add_to_library(model_id=model_id, local_path=str(path), source="local")
 
-    def list_library(self) -> list:
+    def list_library(self) -> list[Dict[str, Any]]:
         """List registered models in the local model library."""
         return list(self._library.values())
 
@@ -124,7 +124,7 @@ def load(
         self,
         model_path: str,
         model_format: Optional[ModelFormat] = None,
-        **kwargs,
+        **kwargs: Any,
     ) -> Dict[str, Any]:
         """
         Load a model and return model objects.
@@ -157,7 +157,7 @@ def load(
         else:
             raise ValueError(f"Unsupported model format: {model_format}")
 
-    def _load_gguf(self, model_path: str, **kwargs) -> Dict[str, Any]:
+    def _load_gguf(self, model_path: str, **kwargs: Any) -> Dict[str, Any]:
         """Load GGUF format model using llama-cpp-python"""
         try:
             from llama_cpp import Llama
@@ -173,7 +173,7 @@ def _load_gguf(self, model_path: str, **kwargs) -> Dict[str, Any]:
 
         return {"model": llm, "tokenizer": None, "format": "gguf"}
 
-    def _load_huggingface(self, model_path: str, **kwargs) -> Dict[str, Any]:
+    def _load_huggingface(self, model_path: str, **kwargs: Any) -> Dict[str, Any]:
         """Load HuggingFace transformers model"""
         try:
             from transformers import AutoModelForCausalLM, AutoTokenizer
@@ -196,7 +196,7 @@ def _load_huggingface(self, model_path: str, **kwargs) -> Dict[str, Any]:
 
         return {"model": model, "tokenizer": tokenizer, "format": "huggingface"}
 
-    def _load_onnx(self, model_path: str, **kwargs) -> Dict[str, Any]:
+    def _load_onnx(self, model_path: str, **kwargs: Any) -> Dict[str, Any]:
         """Load ONNX model"""
         try:
             import onnxruntime as ort
@@ -207,12 +207,12 @@ def _load_onnx(self, model_path: str, **kwargs) -> Dict[str, Any]:
 
         return {"model": session, "tokenizer": None, "format": "onnx"}
 
-    def _load_safetensors(self, model_path: str, **kwargs) -> Dict[str, Any]:
+    def _load_safetensors(self, model_path: str, **kwargs: Any) -> Dict[str, Any]:
         """Load safetensors format model"""
         # Safetensors is typically used with transformers
         return self._load_huggingface(model_path, **kwargs)
 
-    def download(self, model_id: str, **kwargs) -> str:
+    def download(self, model_id: str, **kwargs: Any) -> str:
         """
         Download a model from HuggingFace Hub.
         
@@ -246,7 +246,7 @@ def download(self, model_id: str, **kwargs) -> str:
 
         return local_path
 
-    def list_cached_models(self) -> list:
+    def list_cached_models(self) -> list[str]:
         """List all cached models"""
         return [str(p) for p in self.cache_dir.iterdir() if p.is_dir()]