Dockerfile

# Base image
FROM python:3.12-slim-bookworm AS base

# Set shared environment variables
ENV POETRY_VERSION=1.8.4 \
    POETRY_NO_INTERACTION=1 \
    POETRY_VIRTUALENVS_CREATE=true \
    POETRY_VIRTUALENVS_IN_PROJECT=true \
    POETRY_CACHE_DIR=/tmp/poetry_cache \
    PYTHONDONTWRITEBYTECODE=1 \
    LANG=en_US.UTF-8 \
    LANGUAGE=en_US:en \
    LC_ALL=en_US.UTF-8

# Create user first (HF Spaces requirement)
RUN useradd -m -u 1000 user

# Install system dependencies and set up locales
RUN apt-get update && apt-get install -y \
    curl \
    git \
    gcc \
    python3-dev \
    libgmp-dev \
    libmpfr-dev \
    libmpc-dev \
    nodejs \
    npm \
    postgresql \
    postgresql-contrib \
    locales \
    && rm -rf /var/lib/apt/lists/* \
    && pip install --no-cache-dir "poetry==${POETRY_VERSION}" \
    && sed -i '/en_US.UTF-8/s/^# //g' /etc/locale.gen \
    && locale-gen

# Set up PostgreSQL directories with standard locations
RUN mkdir -p /var/run/postgresql /var/lib/postgresql/data /var/log/postgresql && \
    chown -R user:user /var/run/postgresql /var/lib/postgresql/data /var/log/postgresql && \
    chmod 700 /var/lib/postgresql/data

# Create application directories
RUN mkdir -p /app/api /app/web /data/storage && \
    chown -R user:user /app /data && \
    chmod 777 /data /app

# Switch to user for remaining operations
USER user

# Set environment for user
ENV HOME=/home/user \
    PATH=/usr/lib/postgresql/15/bin:/home/user/.local/bin:$PATH \
    PGDATA=/var/lib/postgresql/data

# Pull official images
FROM langgenius/dify-web:latest AS web
FROM langgenius/dify-api:latest AS api

# Final stage
FROM base

# Set up directory structure
WORKDIR /app
RUN mkdir -p api web /data/storage

# Copy from official images
COPY --from=web --chown=user:user /app/web /app/web/
COPY --from=api --chown=user:user /app/api /app/api/

# Install API dependencies using Poetry
WORKDIR /app/api
COPY --from=api --chown=user /app/api/pyproject.toml /app/api/poetry.lock /app/api/poetry.toml ./
RUN poetry install --no-root --no-dev

# Create symlink for persistent storage
RUN ln -s /data/storage /app/api/storage

# Set environment variables
ENV FLASK_APP=app.py \
    EDITION=SELF_HOSTED \
    DEPLOY_ENV=PRODUCTION \
    MODE=api \
    LOG_LEVEL=INFO \
    DEBUG=false \
    FLASK_DEBUG=false \
    SECRET_KEY=sk-9f73s3ljTXVcMT3Blb3ljTqtsKiGHXVcMT3BlbkFJLK7U \
    CONSOLE_API_URL=https://${SPACE_ID}.hf.space \
    CONSOLE_WEB_URL=https://${SPACE_ID}.hf.space \
    SERVICE_API_URL=https://${SPACE_ID}.hf.space \
    APP_WEB_URL=https://${SPACE_ID}.hf.space \
    DIFY_PORT=7860 \
    DIFY_BIND_ADDRESS=0.0.0.0 \
    DB_USERNAME=user \
    DB_PASSWORD=difyai123456 \
    DB_HOST=localhost \
    DB_PORT=5432 \
    DB_DATABASE=dify \
    PYTHONPATH=/app/api \
    STORAGE_PATH=/data/storage

EXPOSE 7860

# Create startup script
RUN echo '#!/bin/bash\n\
set -e\n\
echo "===== Application Startup at $(date "+%Y-%m-%d %H:%M:%S") ====="\n\
\n\
# Initialize PostgreSQL database if not already initialized\n\
if [ ! -f "$PGDATA/PG_VERSION" ]; then\n\
    echo "Initializing PostgreSQL database..."\n\
    initdb --username=user --pwfile=<(echo "$DB_PASSWORD") --auth=md5 --encoding=UTF8\n\
    \n\
    # Configure PostgreSQL\n\
    echo "local all all trust" > "$PGDATA/pg_hba.conf"\n\
    echo "host all all 127.0.0.1/32 md5" >> "$PGDATA/pg_hba.conf"\n\
    echo "host all all ::1/128 md5" >> "$PGDATA/pg_hba.conf"\n\
    echo "host all all 0.0.0.0/0 md5" >> "$PGDATA/pg_hba.conf"\n\
    \n\
    echo "listen_addresses = '\''*'\''" >> "$PGDATA/postgresql.conf"\n\
    echo "max_connections = 100" >> "$PGDATA/postgresql.conf"\n\
    echo "shared_buffers = 128MB" >> "$PGDATA/postgresql.conf"\n\
fi\n\
\n\
# Start PostgreSQL with detailed logging\n\
echo "Starting PostgreSQL server..."\n\
pg_ctl start -D "$PGDATA" -l /var/log/postgresql/postgresql.log -o "-c logging_collector=on -c log_directory='\''/var/log/postgresql'\'' -c log_filename='\''postgresql-%Y-%m-%d_%H%M%S.log'\'' -c log_statement='\''all'\''" -w\n\
\n\
# Wait for PostgreSQL to start and show logs if there are issues\n\
max_tries=30\n\
count=0\n\
echo "Checking database connection..."\n\
until pg_isready -h localhost -p 5432; do\n\
    if [ $count -eq 0 ]; then\n\
        echo "PostgreSQL logs:"\n\
        tail -n 50 /var/log/postgresql/postgresql.log\n\
    fi\n\
    echo "Waiting for database connection... (${count}/${max_tries})"\n\
    sleep 2\n\
    count=$((count+1))\n\
    if [ $count -gt $max_tries ]; then\n\
        echo "Failed to connect to database after ${max_tries} attempts"\n\
        echo "Last 100 lines of PostgreSQL logs:"\n\
        tail -n 100 /var/log/postgresql/postgresql.log\n\
        exit 1\n\
    fi\n\
done\n\
\n\
# Create database if it doesn'\''t exist\n\
if ! psql -lqt | cut -d \| -f 1 | grep -qw dify; then\n\
    echo "Creating database dify..."\n\
    createdb -U user dify\n\
fi\n\
\n\
echo "Database connection successful"\n\
\n\
# Start application services\n\
cd /app/api && poetry run python -m flask db upgrade\n\
\n\
cd /app/api && poetry run python -m gunicorn app:app \\\n\
    --bind ${DIFY_BIND_ADDRESS:-0.0.0.0}:${DIFY_PORT:-7860} \\\n\
    --worker-class gevent \\\n\
    --workers 1 \\\n\
    --timeout 300 \\\n\
    --preload &\n\
\n\
cd /app/web && node server.js &\n\
\n\
wait' > /app/entrypoint.sh && \
chmod +x /app/entrypoint.sh

WORKDIR /app

CMD ["./entrypoint.sh"]