Dockerfile.jetson

# Start from NVIDIA L4T ML base image
FROM nvcr.io/nvidia/l4t-ml:r32.6.1-py3

# Set environment variables
ENV DEBIAN_FRONTEND=noninteractive \
    PYTHONUNBUFFERED=1 \
    FLASK_ENV=production \
    FLASK_APP=app.py

# Install system dependencies
RUN apt-get update && apt-get install -y --no-install-recommends \
    build-essential \
    libpq-dev \
    libglib2.0-0 \
    libsm6 \
    libxext6 \
    libxrender-dev \
    libgstreamer1.0-0 \
    gstreamer1.0-plugins-base \
    gstreamer1.0-plugins-good \
    gstreamer1.0-plugins-bad \
    gstreamer1.0-plugins-ugly \
    python3-gst-1.0 \
    && rm -rf /var/lib/apt/lists/*

# Create app directory
WORKDIR /app

# Install Python dependencies
COPY requirements.txt .
RUN pip3 install --no-cache-dir -r requirements.txt

# Install additional Jetson-specific packages
RUN pip3 install --no-cache-dir \
    jetson-stats \
    jtop \
    jetson-utils

# Copy application code
COPY . .

# Create necessary directories
RUN mkdir -p /app/logs /app/data /app/models

# Download YOLOv5 model
RUN python3 -c "from ultralytics import YOLO; YOLO('yolov5s.pt')"

# Set up environment for hardware access
ENV OPENBLAS_CORETYPE=ARMV8

# Expose ports
EXPOSE 5000 9090

# Set up entrypoint script
COPY docker-entrypoint.sh /usr/local/bin/
RUN chmod +x /usr/local/bin/docker-entrypoint.sh

# Run the application
ENTRYPOINT ["docker-entrypoint.sh"]
CMD ["gunicorn", "-k", "eventlet", "-w", "4", "-b", "0.0.0.0:5000", "app:app"]