# Opik - Open-source LLM Observability, Evaluation, and Optimization Platform
# https://github.com/comet-ml/opik

x-defaults: &defaults
  restart: unless-stopped
  logging:
    driver: json-file
    options:
      max-size: 100m
      max-file: '3'

services:
  mysql:
    <<: *defaults
    image: ${GLOBAL_REGISTRY:-}mysql:${MYSQL_VERSION:-8.4.2}
    environment:
      TZ: ${TZ:-UTC}
      MYSQL_ROOT_PASSWORD: ${MYSQL_ROOT_PASSWORD:-opik}
      MYSQL_DATABASE: ${MYSQL_DATABASE:-opik}
      MYSQL_USER: ${MYSQL_USER:-opik}
      MYSQL_PASSWORD: ${MYSQL_PASSWORD:-opik}
    volumes:
      - mysql_data:/var/lib/mysql
    healthcheck:
      test: [CMD, mysqladmin, ping, -h, 127.0.0.1, --silent]
      interval: 5s
      timeout: 5s
      retries: 30
      start_period: 30s
    deploy:
      resources:
        limits:
          cpus: ${MYSQL_CPU_LIMIT:-1.0}
          memory: ${MYSQL_MEMORY_LIMIT:-1G}
        reservations:
          cpus: ${MYSQL_CPU_RESERVATION:-0.5}
          memory: ${MYSQL_MEMORY_RESERVATION:-512M}

  redis:
    <<: *defaults
    image: ${GLOBAL_REGISTRY:-}redis:${REDIS_VERSION:-7.2.4-alpine3.19}
    command: redis-server --requirepass ${REDIS_PASSWORD:-opik}
    environment:
      TZ: ${TZ:-UTC}
    volumes:
      - redis_data:/data
    healthcheck:
      test: [CMD, nc, -z, localhost, '6379']
      interval: 5s
      timeout: 5s
      retries: 10
      start_period: 10s
    deploy:
      resources:
        limits:
          cpus: ${REDIS_CPU_LIMIT:-0.5}
          memory: ${REDIS_MEMORY_LIMIT:-512M}
        reservations:
          cpus: ${REDIS_CPU_RESERVATION:-0.25}
          memory: ${REDIS_MEMORY_RESERVATION:-256M}

  zookeeper:
    <<: *defaults
    image: ${GLOBAL_REGISTRY:-}zookeeper:${ZOOKEEPER_VERSION:-3.9.4}
    user: root
    entrypoint:
      - /bin/bash
      - -c
      - |
        mkdir -p /bitnami/zookeeper/data
        chown -R zookeeper:zookeeper /bitnami/zookeeper
        exec gosu zookeeper /docker-entrypoint.sh zkServer.sh start-foreground
    environment:
      TZ: ${TZ:-UTC}
      JVMFLAGS: -Xmx512m
      ZOO_4LW_COMMANDS_WHITELIST: 'srvr,ruok'
      ZOO_DATA_DIR: /bitnami/zookeeper/data
      ZOO_DATA_LOG_DIR: /bitnami/zookeeper/data
    volumes:
      - zookeeper_data:/bitnami/zookeeper
      - zookeeper_data_dir:/data
      - zookeeper_datalog:/datalog
      - zookeeper_logs:/logs
    healthcheck:
      test: [CMD-SHELL, 'echo ruok | nc localhost 2181 | grep -q imok']
      interval: 5s
      timeout: 5s
      retries: 30
      start_period: 10s
    deploy:
      resources:
        limits:
          cpus: ${ZOOKEEPER_CPU_LIMIT:-0.5}
          memory: ${ZOOKEEPER_MEMORY_LIMIT:-1G}
        reservations:
          cpus: ${ZOOKEEPER_CPU_RESERVATION:-0.25}
          memory: ${ZOOKEEPER_MEMORY_RESERVATION:-512M}

  clickhouse-init:
    <<: *defaults
    image: ${GLOBAL_REGISTRY:-}alpine:${ALPINE_VERSION:-latest}
    restart: on-failure
    volumes:
      - ./clickhouse_config:/clickhouse_config_files:ro
      - clickhouse_config:/config
    command: |
      sh -c "
        cp -r /clickhouse_config_files/* /config/ 2>/dev/null || true
        chown -R 1000:1000 /config
      "
    deploy:
      resources:
        limits:
          cpus: '0.1'
          memory: 64M

  clickhouse:
    <<: *defaults
    image: ${GLOBAL_REGISTRY:-}clickhouse/clickhouse-server:${CLICKHOUSE_VERSION:-25.3.6.56-alpine}
    environment:
      TZ: ${TZ:-UTC}
      CLICKHOUSE_DB: ${CLICKHOUSE_DB:-opik}
      CLICKHOUSE_USER: ${CLICKHOUSE_USER:-opik}
      CLICKHOUSE_PASSWORD: ${CLICKHOUSE_PASSWORD:-opik}
      CLICKHOUSE_DEFAULT_ACCESS_MANAGEMENT: 1
    volumes:
      - clickhouse_data:/var/lib/clickhouse
      - clickhouse_logs:/var/log/clickhouse-server
      - clickhouse_config:/etc/clickhouse-server/config.d
    ulimits:
      nofile:
        soft: 262144
        hard: 262144
    depends_on:
      zookeeper:
        condition: service_healthy
      clickhouse-init:
        condition: service_completed_successfully
    healthcheck:
      test: [CMD, wget, --spider, -q, 'http://127.0.0.1:8123/ping']
      interval: 5s
      timeout: 5s
      retries: 30
      start_period: 30s
    deploy:
      resources:
        limits:
          cpus: ${CLICKHOUSE_CPU_LIMIT:-2.0}
          memory: ${CLICKHOUSE_MEMORY_LIMIT:-4G}
        reservations:
          cpus: ${CLICKHOUSE_CPU_RESERVATION:-0.5}
          memory: ${CLICKHOUSE_MEMORY_RESERVATION:-1G}

  minio:
    <<: *defaults
    image: ${GLOBAL_REGISTRY:-}minio/minio:${MINIO_VERSION:-RELEASE.2025-03-12T18-04-18Z}
    environment:
      TZ: ${TZ:-UTC}
      MINIO_ROOT_USER: ${MINIO_ROOT_USER:-opikminio}
      MINIO_ROOT_PASSWORD: ${MINIO_ROOT_PASSWORD:-opikminio123}
    command: server --console-address ":9090" /data
    volumes:
      - minio_data:/data
    ports:
      - '${MINIO_PORT_OVERRIDE:-9000}:9000'
      - '${MINIO_CONSOLE_PORT_OVERRIDE:-9090}:9090'
    healthcheck:
      test: [CMD, curl, -f, 'http://localhost:9000/minio/health/live']
      interval: 5s
      timeout: 10s
      retries: 10
      start_period: 10s
    deploy:
      resources:
        limits:
          cpus: ${MINIO_CPU_LIMIT:-1.0}
          memory: ${MINIO_MEMORY_LIMIT:-1G}
        reservations:
          cpus: ${MINIO_CPU_RESERVATION:-0.25}
          memory: ${MINIO_MEMORY_RESERVATION:-512M}

  minio-init:
    <<: *defaults
    image: ${GLOBAL_REGISTRY:-}minio/mc:${MINIO_MC_VERSION:-RELEASE.2025-03-12T17-29-24Z}
    restart: on-failure
    depends_on:
      minio:
        condition: service_healthy
    entrypoint: >
      /bin/sh -c "
        sleep 5s;
        /usr/bin/mc alias set s3 http://minio:9000 $${MINIO_ROOT_USER:-opikminio} $${MINIO_ROOT_PASSWORD:-opikminio123} --api S3v4;
        /usr/bin/mc mb --ignore-existing s3/public;
        /usr/bin/mc anonymous set download s3/public/;
      "
    deploy:
      resources:
        limits:
          cpus: '0.1'
          memory: 64M

  backend:
    <<: *defaults
    image: ${GLOBAL_REGISTRY:-}ghcr.io/comet-ml/opik/opik-backend:${OPIK_VERSION:-1.10.23}
    command: [bash, -c, './run_db_migrations.sh && ./entrypoint.sh']
    environment:
      TZ: ${TZ:-UTC}
      STATE_DB_PROTOCOL: 'jdbc:mysql://'
      STATE_DB_URL: 'mysql:3306/opik?createDatabaseIfNotExist=true&rewriteBatchedStatements=true'
      STATE_DB_DATABASE_NAME: ${MYSQL_DATABASE:-opik}
      STATE_DB_USER: ${MYSQL_USER:-opik}
      STATE_DB_PASS: ${MYSQL_PASSWORD:-opik}
      ANALYTICS_DB_MIGRATIONS_URL: 'jdbc:clickhouse://clickhouse:8123'
      ANALYTICS_DB_MIGRATIONS_USER: ${CLICKHOUSE_USER:-opik}
      ANALYTICS_DB_MIGRATIONS_PASS: ${CLICKHOUSE_PASSWORD:-opik}
      ANALYTICS_DB_PROTOCOL: HTTP
      ANALYTICS_DB_HOST: clickhouse
      ANALYTICS_DB_PORT: 8123
      ANALYTICS_DB_DATABASE_NAME: ${CLICKHOUSE_DB:-opik}
      ANALYTICS_DB_USERNAME: ${CLICKHOUSE_USER:-opik}
      ANALYTICS_DB_PASS: ${CLICKHOUSE_PASSWORD:-opik}
      JAVA_OPTS: '-Dliquibase.propertySubstitutionEnabled=true -XX:+UseG1GC -XX:MaxRAMPercentage=80.0'
      REDIS_URL: redis://:${REDIS_PASSWORD:-opik}@redis:6379/
      OPIK_OTEL_SDK_ENABLED: 'false'
      OPIK_USAGE_REPORT_ENABLED: ${OPIK_USAGE_REPORT_ENABLED:-true}
      AWS_ACCESS_KEY_ID: ${MINIO_ROOT_USER:-opikminio}
      AWS_SECRET_ACCESS_KEY: ${MINIO_ROOT_PASSWORD:-opikminio123}
      IS_MINIO: 'true'
      S3_URL: http://minio:9000
      PYTHON_EVALUATOR_URL: http://python-backend:8000
      TOGGLE_OPIK_AI_ENABLED: ${TOGGLE_OPIK_AI_ENABLED:-false}
      TOGGLE_GUARDRAILS_ENABLED: ${TOGGLE_GUARDRAILS_ENABLED:-false}
      TOGGLE_WELCOME_WIZARD_ENABLED: ${TOGGLE_WELCOME_WIZARD_ENABLED:-true}
      CORS: ${CORS:-false}
      ATTACHMENTS_STRIP_MIN_SIZE: ${ATTACHMENTS_STRIP_MIN_SIZE:-256000}
      JACKSON_MAX_STRING_LENGTH: ${JACKSON_MAX_STRING_LENGTH:-104857600}
    ports:
      - '${OPIK_BACKEND_PORT_OVERRIDE:-3003}:3003'
    depends_on:
      mysql:
        condition: service_healthy
      clickhouse:
        condition: service_healthy
      minio:
        condition: service_healthy
      redis:
        condition: service_healthy
    healthcheck:
      test: [CMD, curl, -f, 'http://localhost:8080/health-check']
      interval: 10s
      timeout: 10s
      retries: 30
      start_period: 60s
    deploy:
      resources:
        limits:
          cpus: ${BACKEND_CPU_LIMIT:-2.0}
          memory: ${BACKEND_MEMORY_LIMIT:-2G}
        reservations:
          cpus: ${BACKEND_CPU_RESERVATION:-0.5}
          memory: ${BACKEND_MEMORY_RESERVATION:-1G}
    volumes:
      - backend_tmp:/tmp

  python-backend:
    <<: *defaults
    image: ${GLOBAL_REGISTRY:-}ghcr.io/comet-ml/opik/opik-python-backend:${OPIK_VERSION:-1.10.23}
    privileged: true
    environment:
      TZ: ${TZ:-UTC}
      OPIK_OTEL_SDK_ENABLED: 'false'
      PYTHON_CODE_EXECUTOR_IMAGE_TAG: ${OPIK_VERSION:-1.10.23}
      PYTHON_CODE_EXECUTOR_STRATEGY: ${PYTHON_CODE_EXECUTOR_STRATEGY:-process}
      PYTHON_CODE_EXECUTOR_CONTAINERS_NUM: ${PYTHON_CODE_EXECUTOR_CONTAINERS_NUM:-5}
      PYTHON_CODE_EXECUTOR_EXEC_TIMEOUT_IN_SECS: ${PYTHON_CODE_EXECUTOR_EXEC_TIMEOUT_IN_SECS:-3}
      PYTHON_CODE_EXECUTOR_ALLOW_NETWORK: ${PYTHON_CODE_EXECUTOR_ALLOW_NETWORK:-false}
      PYTHON_CODE_EXECUTOR_CPU_SHARES: ${PYTHON_CODE_EXECUTOR_CPU_SHARES:-512}
      PYTHON_CODE_EXECUTOR_MEM_LIMIT: ${PYTHON_CODE_EXECUTOR_MEM_LIMIT:-256m}
      OPIK_VERSION: ${OPIK_VERSION:-1.10.23}
      OPIK_REVERSE_PROXY_URL: http://frontend:5173/api
      PYTHON_BACKEND_PORT: ${PYTHON_BACKEND_PORT:-8000}
      OPENAI_API_KEY: ${OPENAI_API_KEY:-}
      ANTHROPIC_API_KEY: ${ANTHROPIC_API_KEY:-}
      OPENROUTER_API_KEY: ${OPENROUTER_API_KEY:-}
      REDIS_URL: redis://:${REDIS_PASSWORD:-opik}@redis:6379/0
      RQ_WORKER_ENABLED: ${RQ_WORKER_ENABLED:-true}
      OPTSTUDIO_MAX_CONCURRENT_JOBS: ${OPTSTUDIO_MAX_CONCURRENT_JOBS:-5}
      OPTSTUDIO_LOG_LEVEL: ${OPTSTUDIO_LOG_LEVEL:-INFO}
      OPTSTUDIO_LLM_MAX_TOKENS: ${OPTSTUDIO_LLM_MAX_TOKENS:-8192}
      RQ_WORKER_TTL_FAILURE: ${RQ_WORKER_TTL_FAILURE:-86400}
      OPIK_URL_OVERRIDE: http://backend:8080
    depends_on:
      backend:
        condition: service_healthy
      redis:
        condition: service_healthy
    healthcheck:
      test: [CMD, wget, --spider, -q, 'http://127.0.0.1:8000/healthcheck']
      interval: 10s
      timeout: 10s
      retries: 10
      start_period: 30s
    deploy:
      resources:
        limits:
          cpus: ${PYTHON_BACKEND_CPU_LIMIT:-1.0}
          memory: ${PYTHON_BACKEND_MEMORY_LIMIT:-1G}
        reservations:
          cpus: ${PYTHON_BACKEND_CPU_RESERVATION:-0.5}
          memory: ${PYTHON_BACKEND_MEMORY_RESERVATION:-512M}
    volumes:
      - python_backend_docker:/var/lib/docker

  frontend:
    <<: *defaults
    image: ${GLOBAL_REGISTRY:-}ghcr.io/comet-ml/opik/opik-frontend:${OPIK_VERSION:-1.10.23}
    ports:
      - '${OPIK_PORT_OVERRIDE:-5173}:80'
    environment:
      TZ: ${TZ:-UTC}
      NGINX_PORT: ${NGINX_PORT:-80}
      NGINX_CONF_SUFFIX: ${NGINX_CONF_SUFFIX:-local}
    depends_on:
      backend:
        condition: service_healthy
    healthcheck:
      test: [CMD-SHELL, 'curl --fail --fail-early http://localhost:80/']
      interval: 10s
      timeout: 10s
      retries: 10
      start_period: 10s
    deploy:
      resources:
        limits:
          cpus: ${FRONTEND_CPU_LIMIT:-0.5}
          memory: ${FRONTEND_MEMORY_LIMIT:-512M}
        reservations:
          cpus: ${FRONTEND_CPU_RESERVATION:-0.25}
          memory: ${FRONTEND_MEMORY_RESERVATION:-256M}

volumes:
  mysql_data:
  redis_data:
  zookeeper_data:
  zookeeper_data_dir:
  zookeeper_datalog:
  zookeeper_logs:
  clickhouse_data:
  clickhouse_logs:
  clickhouse_config:
  minio_data:
  backend_tmp:
  python_backend_docker: