Fix IMS weather staleness, add data freshness endpoint

backend/Dockerfile

@@ -0,0 +1,28 @@
+FROM python:3.12-slim
+
+WORKDIR /app
+
+# Install dependencies first (layer caching)
+COPY requirements.txt .
+COPY backend/requirements.txt backend/
+RUN pip install --no-cache-dir -r requirements.txt -r backend/requirements.txt
+
+# Non-root user for security
+RUN groupadd -r solarwine && useradd -r -g solarwine solarwine
+
+# Copy application code and data cache
+COPY src/ src/
+COPY config/ config/
+COPY backend/ backend/
+COPY Data/ Data/
+
+ENV PYTHONPATH=/app
+
+# Switch to non-root
+USER solarwine
+
+# HuggingFace Spaces requires port 7860
+EXPOSE 7860
+HEALTHCHECK --interval=30s --timeout=5s --retries=3 \
+    CMD python -c "import urllib.request; urllib.request.urlopen('http://localhost:7860/api/health')" || exit 1
+CMD ["uvicorn", "backend.api.main:app", "--host", "0.0.0.0", "--port", "7860"]

backend/HF_README.md

@@ -0,0 +1,26 @@
+---
+title: SolarWine API
+emoji: 🌿
+colorFrom: green
+colorTo: yellow
+sdk: docker
+app_port: 7860
+private: true
+---
+
+# SolarWine API
+
+FastAPI backend for the SolarWine agrivoltaic vineyard control system.
+
+## Endpoints
+
+- `GET /api/health` — health check
+- `GET /api/weather/current` — current weather (IMS station 43)
+- `GET /api/sensors/snapshot` — vine sensor readings (ThingsBoard)
+- `GET /api/energy/current` — current power output
+- `GET /api/photosynthesis/current` — photosynthesis rate (FvCB/ML)
+- `GET /api/control/status` — last control loop tick
+- `POST /api/chatbot/message` — AI vineyard advisor
+- `GET /api/biology/rules` — biology rules
+
+Interactive docs at `/docs`.

backend/api/main.py

@@ -74,12 +74,12 @@ def _check_data_paths():
             log.warning("Missing %s — %s", rel_path, msg)
 
 
-async def _ims_refresh_loop(interval_sec: int = 6 * 3600):
-    """Background loop: refresh the IMS weather cache every `interval_sec` seconds.
+async def _ims_refresh_loop(interval_sec: int = 2 * 3600):
+    """Background loop: refresh the IMS weather cache every 2 hours.
 
-    Fetches the last 14 days of IMS data, overwrites the local CSV cache,
-    and invalidates the in-memory WeatherService cache so the next request
-    picks up fresh data.
+    Fetches the last 7 days of IMS data in 3-day chunks (IMS API is
+    unreliable with large ranges), overwrites the local CSV cache,
+    and invalidates the in-memory WeatherService cache.
     """
     import asyncio
     from backend.api.events import event_bus
@@ -89,17 +89,19 @@ async def _ims_refresh_loop(interval_sec: int = 6 * 3600):
             from datetime import date, timedelta
             from src.data.ims_client import IMSClient
             end = date.today()
-            start = end - timedelta(days=14)
+            start = end - timedelta(days=7)
             client = IMSClient()
             loop = asyncio.get_event_loop()
             df = await loop.run_in_executor(
                 None,
                 lambda: client.fetch_and_cache(
-                    str(start), str(end), chunk_days=14
+                    str(start), str(end), chunk_days=3
                 ),
             )
             rows = len(df) if df is not None and not df.empty else 0
             log.info("IMS cache refreshed: %d rows (%s → %s)", rows, start, end)
+            if rows == 0:
+                log.warning("IMS refresh returned 0 rows — API may be down")
             # Invalidate WeatherService in-memory cache so next request uses fresh data
             try:
                 from backend.api.deps import get_datahub
@@ -107,9 +109,10 @@ async def _ims_refresh_loop(interval_sec: int = 6 * 3600):
                 hub.weather._df_cache._store.clear()
             except Exception:
                 pass
-            await event_bus.notify("weather")
+            if rows > 0:
+                await event_bus.notify("weather")
         except Exception as exc:
-            log.error("IMS refresh failed: %s", exc)
+            log.error("IMS refresh failed: %s", exc, exc_info=True)
         await asyncio.sleep(interval_sec)
 
 

backend/api/routes/health.py

@@ -1,13 +1,14 @@
-"""Health check endpoint."""
+"""Health check endpoints."""
 
 from __future__ import annotations
 
 import asyncio
 import os
 
-from fastapi import APIRouter
+from fastapi import APIRouter, Depends
 
-from backend.api.deps import get_redis_client
+from backend.api.deps import get_datahub, get_redis_client
+from src.data.data_providers import DataHub
 
 router = APIRouter()
 
@@ -47,3 +48,50 @@ async def health():
         "ims_configured": bool(os.environ.get("IMS_API_TOKEN")),
         "gemini_configured": bool(os.environ.get("GOOGLE_API_KEY")),
     }
+
+
+@router.get("/health/data")
+async def health_data(hub: DataHub = Depends(get_datahub)):
+    """Data freshness check — shows age of each cached data source."""
+    sources = {}
+
+    # Weather
+    try:
+        wx = hub.weather.get_current()
+        if wx and "error" not in wx:
+            sources["weather"] = {
+                "age_minutes": round(float(wx.get("age_minutes", -1)), 1),
+                "last_reading": wx.get("timestamp_local"),
+                "ok": float(wx.get("age_minutes", 9999)) < 120,
+            }
+        else:
+            sources["weather"] = {"ok": False, "error": wx.get("error", "unavailable")}
+    except Exception as exc:
+        sources["weather"] = {"ok": False, "error": str(exc)}
+
+    # Sensors
+    try:
+        snap = hub.vine_sensors.get_snapshot(light=True)
+        if snap and "error" not in snap:
+            stale = snap.get("staleness_minutes", -1)
+            sources["sensors"] = {
+                "age_minutes": round(float(stale), 1) if stale is not None else None,
+                "ok": stale is not None and float(stale) < 30,
+            }
+        else:
+            sources["sensors"] = {"ok": False, "error": snap.get("error", "unavailable")}
+    except Exception as exc:
+        sources["sensors"] = {"ok": False, "error": str(exc)}
+
+    # Energy
+    try:
+        en = hub.energy.get_current()
+        if en and "error" not in en:
+            sources["energy"] = {"ok": True, "power_kw": en.get("power_kw")}
+        else:
+            sources["energy"] = {"ok": False, "error": en.get("error", "unavailable")}
+    except Exception as exc:
+        sources["energy"] = {"ok": False, "error": str(exc)}
+
+    all_ok = all(s.get("ok", False) for s in sources.values())
+    return {"status": "ok" if all_ok else "degraded", "sources": sources}

backend/requirements.txt

@@ -1,7 +1,7 @@
-# Backend-specific dependencies
+# Backend-specific dependencies (on top of root requirements.txt)
 fastapi>=0.115.0
 uvicorn[standard]>=0.34.0
 pydantic>=2.0
 slowapi>=0.1.9
 PyJWT>=2.8.0
-sentry-sdk[fastapi]>=2.0
+sentry-sdk[fastapi]>=2.0  # optional: set SENTRY_DSN to enable

requirements.txt

@@ -1,10 +1,18 @@
-# API-only dependencies (slim — no torch/chronos/streamlit)
-pandas>=2.0
-numpy>=1.26
-scikit-learn>=1.3
-requests>=2.31
-python-dotenv>=1.0
+# Photosynthesis Prediction Model - dependencies
+# Install: pip install -r requirements.txt
+
+pandas==2.3.3
+numpy==2.4.2
+scikit-learn==1.8.0
+matplotlib==3.10.8
+seaborn==0.13.2
+requests==2.32.5
+python-dotenv==1.2.1
+streamlit==1.54.0
+plotly==6.5.2
 xgboost>=2.0
 pvlib>=0.10.0
 astral>=3.2
+chronos-forecasting>=2.0
+torch>=2.0
 google-genai>=1.0