docker-compose-apache.yml

name: hms

# ─── Verbose console logging ──────────────────────────────────────────────────
# conf/hive-log4j2.properties (bind-mounted below) defines a Console/SYSTEM_OUT
# appender named CONSOLE and sets the root logger to DEBUG.  This file fully
# replaces the image's baked-in log4j2 config, so all HMS log output goes to
# stdout and is visible in `docker compose logs -f metastore`.
#
# Hive 3.x images do not support the VERBOSE env var, so it is omitted.
# ──────────────────────────────────────────────────────────────────────────────

services:

  # ── PostgreSQL ───────────────────────────────────────────────────────────────
  postgres:
    image: postgres:15
    container_name: postgres
    hostname: postgres
    restart: unless-stopped
    environment:
      POSTGRES_DB: metastore_db
      POSTGRES_USER: hive
      POSTGRES_PASSWORD: password
      POSTGRES_INITDB_ARGS: "--auth-host=md5 --auth-local=md5"
    command: ["postgres", "-c", "password_encryption=md5"]
    ports:
      - "5432:5432"
    volumes:
      - hive-db:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U hive -d metastore_db"]
      interval: 10s
      timeout: 5s
      retries: 5
      start_period: 15s
    networks:
      - hive

  # ── Download PostgreSQL JDBC driver (one-shot, skipped if already present) ──
  # The HMS entrypoint copies every *.jar found in /tmp/ext-jars into
  # $HIVE_HOME/lib before starting the service (see entrypoint.sh lines 28-35).
#  postgres-jar:
#    image: alpine/curl
#    container_name: postgres-jar
#    volumes:
#      - hms-jars:/downloads
#    command: >-
#      sh -c "[ -f /downloads/postgresql.jar ] ||
#             curl -fSL -o /downloads/postgresql.jar
#             https://jdbc.postgresql.org/download/postgresql-42.7.3.jar"

  # ── Hive Metastore (HMS) ─────────────────────────────────────────────────────
  metastore:
    image: apache/hive:${HIVE_VERSION:-3.1.3}
    container_name: metastore
    hostname: metastore
    restart: unless-stopped
    depends_on:
      postgres:
        condition: service_healthy
#      postgres-jar:
#        condition: service_completed_successfully
    environment:
      DB_DRIVER: postgres
      SERVICE_NAME: metastore

      # ── PostgreSQL JDBC connection ─────────────────────────────────────────
      # -Xmx1G is kept here; entrypoint.sh prepends it again via
      # HADOOP_CLIENT_OPTS so the value here sets the effective heap.
      SERVICE_OPTS: >-
        -Xmx1G
        -Djavax.jdo.option.ConnectionDriverName=org.postgresql.Driver
        -Djavax.jdo.option.ConnectionURL=jdbc:postgresql://postgres:5432/metastore_db
        -Djavax.jdo.option.ConnectionUserName=hive
        -Djavax.jdo.option.ConnectionPassword=password

      # Schema will be initialised on first start; set to 'true' to skip.
      IS_RESUME: "false"

      HADOOP_CLASSPATH: /opt/hadoop/share/hadoop/tools/lib/*

    ports:
      - "9083:9083"
    volumes:
      - warehouse:/opt/hive/data/warehouse
      # Jars placed here are auto-copied to $HIVE_HOME/lib by the entrypoint.
      - hms-jars:/tmp/ext-jars:ro
      # Custom log4j2 config: console-only appender at DEBUG level.
      # Fully replaces the image's baked-in hive-log4j2.properties.
      - ./conf/hive-log4j2.properties:/opt/hive/conf/hive-log4j2.properties:ro
    networks:
      - hive

volumes:
  hive-db:
  warehouse:
  hms-jars:

networks:
  hive:
    name: hive