docker-compose.yml

version: '3'

services:
  # Zookeeper service for Kafka coordination and configuration management
  zookeeper:
      image: confluentinc/cp-zookeeper:7.4.0
      hostname: zookeeper
      container_name: zookeeper
      ports:
        - "2182:2182"  
      environment:
        ZOOKEEPER_CLIENT_PORT: 2182
        ZOOKEEPER_TICK_TIME: 2000  
      healthcheck:
        test: ['CMD', 'bash', '-c', "echo 'ruok' | nc localhost 2182"]  
        interval: 10s
        timeout: 5s
        retries: 5
      networks:
        - confluent  # Part of the confluent network

  # Kafka broker service for managing Kafka topics and message streams
  broker:
    image: confluentinc/cp-server:7.4.0
    hostname: broker
    container_name: broker
    depends_on:
      zookeeper:
        condition: service_healthy  # Waits for Zookeeper to be healthy before starting
    ports:
      - "9092:9092"  # External port for client connections
      - "9101:9101"  # JMX port for monitoring
    environment:
      KAFKA_BROKER_ID: 1
      KAFKA_ZOOKEEPER_CONNECT: 'zookeeper:2182'  # Connection to Zookeeper
      KAFKA_LISTENER_SECURITY_PROTOCOL_MAP: PLAINTEXT:PLAINTEXT,PLAINTEXT_HOST:PLAINTEXT
      KAFKA_ADVERTISED_LISTENERS: PLAINTEXT://broker:29092,PLAINTEXT_HOST://localhost:9092
      KAFKA_OFFSETS_TOPIC_REPLICATION_FACTOR: 1  # Replication factor for internal topics
    networks:
      - confluent
    healthcheck:
      test: [ "CMD", "bash", "-c", 'nc -z localhost 9092' ]  # Check if Kafka is accessible on port 9092
      interval: 10s
      timeout: 5s
      retries: 5

  # Schema Registry service for managing schemas for Kafka data
  schema-registry:
    image: confluentinc/cp-schema-registry:7.4.0
    hostname: schema-registry
    container_name: schema-registry
    depends_on:
      broker:
        condition: service_healthy  # Depends on Kafka broker
    ports:
      - "8081:8081"  # Schema Registry API port
    environment:
      SCHEMA_REGISTRY_KAFKASTORE_BOOTSTRAP_SERVERS: 'broker:29092'
      SCHEMA_REGISTRY_LISTENERS: http://0.0.0.0:8081  # Exposing Schema Registry on port 8081
    networks:
      - confluent
    healthcheck:
      test: [ "CMD", "curl", "-f", "http://localhost:8081/" ]  # Health check using a curl request
      interval: 30s
      timeout: 10s
      retries: 5

  # Control Center for monitoring and managing the Kafka cluster
  control-center:
    image: confluentinc/cp-enterprise-control-center:7.4.0
    hostname: control-center
    container_name: control-center
    depends_on:
      broker:
        condition: service_healthy
      schema-registry:
        condition: service_healthy
    ports:
      - "9021:9021"  # Web UI port for the Control Center
    environment:
      CONTROL_CENTER_BOOTSTRAP_SERVERS: 'broker:29092'
      CONTROL_CENTER_SCHEMA_REGISTRY_URL: "http://schema-registry:8081"
    networks:
      - confluent
    healthcheck:
      test: [ "CMD", "curl", "-f", "http://localhost:9021/health" ]  
      interval: 30s
      timeout: 10s
      retries: 5

  # Airflow webserver for orchestrating data pipelines
  webserver:
    image: apache/airflow:2.6.0-python3.9
    command: webserver
    entrypoint: ['/opt/airflow/script/entrypoint.sh']
    depends_on:
      - postgres  # Depends on PostgreSQL database for metadata storage
    environment:
      - AIRFLOW__DATABASE__SQL_ALCHEMY_CONN=postgresql+psycopg2://airflow:airflow@postgres:5432/airflow
    volumes:
      - ./dags:/opt/airflow/dags  # Mounts local DAGs folder into the container
      - ./script/entrypoint.sh:/opt/airflow/script/entrypoint.sh  # Custom entrypoint script
      - ./requirements.txt:/opt/airflow/requirements.txt  # Python dependencies
    ports:
      - "8088:8080"  # Airflow webserver port
    healthcheck:
      test: ['CMD-SHELL', "[ -f /opt/airflow/airflow-webserver.pid ]"]  
      interval: 30s
      timeout: 30s
      retries: 3
    networks:
      - confluent

  # PostgreSQL database service for Airflow
  postgres:
    image: postgres:14.0
    environment:
      - POSTGRES_USER=airflow
      - POSTGRES_PASSWORD=airflow
      - POSTGRES_DB=airflow
    networks:
      - confluent

  # Airflow scheduler for executing tasks according to DAG definitions
  scheduler:
    image: apache/airflow:2.6.0-python3.9
    depends_on:
      webserver:
        condition: service_healthy
    volumes:
      - ./dags:/opt/airflow/dags
      - ./script/entrypoint.sh:/opt/airflow/script/entrypoint.sh
      - ./requirements.txt:/opt/airflow/requirements.txt
    environment:
      - AIRFLOW__DATABASE__SQL_ALCHEMY_CONN=postgresql+psycopg2://airflow:airflow@postgres:5432/airflow
    command: bash -c "pip install -r ./requirements.txt && airflow db upgrade && airflow scheduler"  # Initializes DB and starts scheduler
    networks:
      - confluent

  # Spark master node for distributed data processing
  spark-master:
    image: bitnami/spark:latest
    command: bin/spark-class org.apache.spark.deploy.master.Master
    ports:
      - "9090:8080"  # Web UI for the Spark master
      - "7077:7077"  # Spark master port for worker connections
    networks:
      - confluent

  # Spark worker node for distributed data processing
  spark-worker:
    image: bitnami/spark:latest
    command: bin/spark-class org.apache.spark.deploy.worker.Worker spark://spark-master:7077
    depends_on:
      - spark-master  # Depends on Spark master being healthy
    environment:
      SPARK_MODE: worker
      SPARK_WORKER_CORES: 2
      SPARK_WORKER_MEMORY: 1g
    networks:
      - confluent

  # Cassandra database for storing processed data
  cassandra_db:
    image: cassandra:latest
    container_name: cassandra
    hostname: cassandra
    ports:
      - "9042:9042"  
    environment:
      - MAX_HEAP_SIZE=512M
      - HEAP_NEWSIZE=100M
      - CASSANDRA_USERNAME=cassandra
      - CASSANDRA_PASSWORD=cassandra
    networks:
      - confluent    

networks:
  confluent: