Dockerfile

# ── Stage 1: Kafka binaries ──────────────────────────────────────────────────
FROM apache/kafka:3.9.0 AS kafka-src

# ── Stage 2: Frontend build ───────────────────────────────────────────────────
FROM node:20-alpine AS frontend-build
WORKDIR /build
COPY services/frontend/package.json services/frontend/package-lock.json ./
RUN npm install --prefer-offline
COPY services/frontend .
RUN npm run build

# ── Stage 3: Main runtime ─────────────────────────────────────────────────────
FROM python:3.12-slim

# System packages
RUN apt-get update && apt-get install -y --no-install-recommends \
    openjdk-21-jre-headless \
    nginx \
    supervisor \
    postgresql \
    redis-server \
    procps \
    && rm -rf /var/lib/apt/lists/*

# Kafka JARs + scripts from official image (JVM-based, no native binary issues)
COPY --from=kafka-src /opt/kafka /opt/kafka
ENV PATH="/opt/kafka/bin:$PATH"

# Java home for PySpark — symlink resolves across amd64/arm64
RUN ln -s "$(dirname $(dirname $(readlink -f $(which java))))" /opt/java
ENV JAVA_HOME=/opt/java

# Python packages
RUN pip install --no-cache-dir \
    "fastapi>=0.111" \
    "uvicorn[standard]" \
    asyncpg \
    "redis[asyncio]>=5.0" \
    "kafka-python-ng>=2.2" \
    prometheus-client \
    "pydantic>=2.7" \
    "mlflow>=2.13" \
    psycopg2-binary \
    "pyspark==3.5.3" \
    "numpy>=1.26" \
    "httpx>=0.27"

# Pre-download Spark Kafka connector JAR so streaming.py starts offline
RUN python -c "\
from pyspark.sql import SparkSession; \
SparkSession.builder \
    .config('spark.jars.packages', 'org.apache.spark:spark-sql-kafka-0-10_2.12:3.5.3') \
    .config('spark.ui.enabled', 'false') \
    .getOrCreate().stop()"

# Frontend static files
COPY --from=frontend-build /build/dist /app/frontend/dist

# App sources
COPY services/api        /app/api
COPY services/spark-jobs /app/spark-jobs
COPY services/event-simulator /app/event-simulator

# HF infra configs + scripts
COPY infra/hf /app/infra/hf
RUN chmod +x /app/infra/hf/*.sh

# nginx: replace default config
RUN rm -f /etc/nginx/sites-enabled/default
COPY infra/hf/nginx.conf /etc/nginx/conf.d/cinematch.conf

# Pre-seeded bootstrap data (committed to repo after local `make bootstrap`)
COPY data /app/data

# Runtime dirs
RUN mkdir -p /var/lib/postgresql/data \
             /var/lib/kafka-logs \
             /var/log/supervisor \
             /var/log/kafka \
    && chown postgres:postgres /var/lib/postgresql/data

WORKDIR /app
EXPOSE 7860
CMD ["/app/infra/hf/init.sh"]