docker-compose.yaml

version: "3.9"

services:
  crawler:
    image: bronbro/spacebox-crawler:latest
    environment:
      ENV_FILE: "/opt/.env"
    restart: unless-stopped
    volumes:
      - .env:/opt/.env
    ports:
      - '2112:2112'
    logging:
      driver: "json-file"
      options:
        max-size: "300m"
        max-file: "5"
    depends_on:
      kafka:
        condition: service_healthy
      zookeeper:
        condition: service_started
      mongo-crawler:
        condition: service_started
  kafka-ui:
    container_name: spacebox-kafka-ui
    image: provectuslabs/kafka-ui:latest
    ports:
      - '8080:8080'
    restart: unless-stopped
    depends_on:
      kafka:
        condition: service_started
      zookeeper:
        condition: service_started
    environment:
      KAFKA_CLUSTERS_0_NAME: local
      KAFKA_CLUSTERS_0_BOOTSTRAPSERVERS: kafka:9093
      AUTH_TYPE: "LOGIN_FORM"
      SPRING_SECURITY_USER_NAME: default
      SPRING_SECURITY_USER_PASSWORD: ${KAFKA_UI_PASSWORD:-password}
  zookeeper:
    image: bitnami/zookeeper:3.9.2
    ports:
      - '2181:2181'
    environment:
      - ALLOW_ANONYMOUS_LOGIN=yes
    restart: unless-stopped
    container_name: spacebox-zookeeper
    volumes:
      - ./volumes/zookeeper:/bitnami/zookeeper
  kafka:
    image: bitnami/kafka:3.7.0
    ports:
      - '9092:9092'
      - '9093:9093'
    hostname: kafka
    environment:
      - KAFKA_CFG_ZOOKEEPER_CONNECT=zookeeper:2181
      - ALLOW_PLAINTEXT_LISTENER=yes
      - KAFKA_CFG_LISTENER_SECURITY_PROTOCOL_MAP=CLIENT:PLAINTEXT,EXTERNAL:PLAINTEXT
      - KAFKA_CFG_LISTENERS=CLIENT://:9093,EXTERNAL://:9092
      - KAFKA_CFG_ADVERTISED_LISTENERS=CLIENT://kafka:9093,EXTERNAL://localhost:9092
      - KAFKA_CFG_INTER_BROKER_LISTENER_NAME=CLIENT
      - KAFKA_CFG_DELETE_TOPIC_ENABLE=true
      - KAFKA_CFG_LOG_RETENTION_BYTES=16106127360
      - KAFKA_CFG_MESSAGE_MAX_BYTES=${MAX_MESSAGE_MAX_BYTES}
      - KAFKA_CFG_REPLICA_FETCH_MAX_BYTES=157286400
    volumes:
      - ./volumes/kafka:/bitnami/kafka
    restart: unless-stopped
    depends_on:
      zookeeper:
        condition: service_started
    healthcheck:
      test: kafka-topics.sh --list --bootstrap-server localhost:9092
      interval: 120s
      timeout: 60s
      retries: 5
  migration:
    image: migrate/migrate:latest
    command: |
      -path=/migrations/clickhouse 
      -database="clickhouse://clickhouse:9000/spacebox?username=default&x-multi-statement=true&password=${CLICKHOUSE_PASSWORD:-password}" 
      up
    volumes:
      - ./migrations:/migrations
    depends_on:
      clickhouse:
        condition: service_healthy
  clickhouse:
    image: clickhouse/clickhouse-server:24.5.1
    restart: unless-stopped
    hostname: clickhouse
    ulimits:
      nofile:
        soft: 262144
        hard: 262144
    ports:
      - '8123:8123'
      - '9000:9000'
      - '9009:9009'
      - '8001:8001'
    environment:
      CLICKHOUSE_INIT_TIMEOUT: 60
      CLICKHOUSE_DB: spacebox
      CLICKHOUSE_USER: default
      CLICKHOUSE_PASSWORD: ${CLICKHOUSE_PASSWORD:-password}
    healthcheck:
      test: wget --no-verbose --tries=1 --spider http://localhost:8123/ping
      interval: 60s
      timeout: 10s
      retries: 5
    volumes:
      - ./volumes/clickhouse/data:/var/lib/clickhouse
      - ./volumes/clickhouse/logs:/var/log/clickhouse-server
      - ./config/clickhouse/users.d/:/etc/clickhouse-server/users.d/
      - ./config/clickhouse/config.d/:/etc/clickhouse-server/config.d/
  mongo-crawler:
    image: mongo:6.0.3
    restart: always
    hostname: mongo-1
    environment:
      MONGO_INITDB_ROOT_USERNAME: spacebox_user
      MONGO_INITDB_ROOT_PASSWORD: spacebox_password
      MONGO_INITDB_DATABASE: spacebox
    ports:
      - '27018:27017'
    volumes:
      - ./volumes/mongo/crawler:/data/db