CHORUS/docker/docker-compose.yml

version: "3.9"

services:
  chorus:
    image: localhost:5000/chorus:march8-evidence-20260226-2
    environment:
      - CHORUS_LICENSE_ID_FILE=/run/secrets/chorus_license_id
      - CHORUS_CLUSTER_ID=${CHORUS_CLUSTER_ID:-docker-cluster}
      - CHORUS_KACHING_URL=${CHORUS_KACHING_URL:-http://host.docker.internal:8099}
      - CHORUS_AGENT_ID=${CHORUS_AGENT_ID:-}
      - CHORUS_SPECIALIZATION=${CHORUS_SPECIALIZATION:-general_developer}
      - CHORUS_MAX_TASKS=${CHORUS_MAX_TASKS:-3}
      - CHORUS_CAPABILITIES=general_development,task_coordination,admin_election
      - CHORUS_API_PORT=8080
      - CHORUS_HEALTH_PORT=8081
      - CHORUS_P2P_PORT=9000
      - CHORUS_BIND_ADDRESS=0.0.0.0
      - CHORUS_MDNS_ENABLED=false
      - CHORUS_DIALS_PER_SEC=5
      - CHORUS_MAX_CONCURRENT_DHT=16
      - CHORUS_ELECTION_MIN_TERM=120s
      - CHORUS_LEADER_MIN_TERM=240s
      - ASSIGN_URL=${ASSIGN_URL:-}
      - TASK_SLOT=${TASK_SLOT:-}
      - TASK_ID=${TASK_ID:-}
      - NODE_ID=${NODE_ID:-}
      - WHOOSH_API_BASE_URL=${SWOOSH_API_BASE_URL:-http://swoosh:8080}
      - WHOOSH_API_ENABLED=true
      - BOOTSTRAP_JSON=/config/bootstrap.json
      - CHORUS_BOOTSTRAP_PEERS=${CHORUS_BOOTSTRAP_PEERS:-}
      - CHORUS_AI_PROVIDER=${CHORUS_AI_PROVIDER:-resetdata}
      - RESETDATA_BASE_URL=${RESETDATA_BASE_URL:-https://app.resetdata.ai/api/v1}
      - RESETDATA_API_KEY_FILE=/run/secrets/resetdata_api_key
      - RESETDATA_MODEL=${RESETDATA_MODEL:-openai/gpt-oss-120b}
      - OLLAMA_ENDPOINT=${OLLAMA_ENDPOINT:-http://host.docker.internal:11434}
      - CHORUS_MODELS=${CHORUS_MODELS:-meta/llama-3.1-8b-instruct}
      - CHORUS_DEFAULT_REASONING_MODEL=${CHORUS_DEFAULT_REASONING_MODEL:-meta/llama-3.1-8b-instruct}
      - CHORUS_LIGHTRAG_ENABLED=${CHORUS_LIGHTRAG_ENABLED:-true}
      - CHORUS_LIGHTRAG_BASE_URL=${CHORUS_LIGHTRAG_BASE_URL:-http://host.docker.internal:9621}
      - CHORUS_LIGHTRAG_TIMEOUT=${CHORUS_LIGHTRAG_TIMEOUT:-30s}
      - CHORUS_LIGHTRAG_API_KEY=${CHORUS_LIGHTRAG_API_KEY:-your-secure-api-key-here}
      - CHORUS_LIGHTRAG_DEFAULT_MODE=${CHORUS_LIGHTRAG_DEFAULT_MODE:-hybrid}
      - LOG_LEVEL=${LOG_LEVEL:-info}
      - LOG_FORMAT=${LOG_FORMAT:-structured}
      - CHORUS_BACKBEAT_ENABLED=${CHORUS_BACKBEAT_ENABLED:-true}
      - CHORUS_BACKBEAT_CLUSTER_ID=${CHORUS_BACKBEAT_CLUSTER_ID:-chorus-production}
      - CHORUS_BACKBEAT_AGENT_ID=${CHORUS_BACKBEAT_AGENT_ID:-}
      - CHORUS_BACKBEAT_NATS_URL=${CHORUS_BACKBEAT_NATS_URL:-nats://backbeat-nats:4222}
      - CHORUS_TRANSPORT_TELEMETRY_INTERVAL=${CHORUS_TRANSPORT_TELEMETRY_INTERVAL:-30s}
      - CHORUS_TRANSPORT_TELEMETRY_SUBJECT=${CHORUS_TRANSPORT_TELEMETRY_SUBJECT:-chorus.telemetry.transport}
      - CHORUS_TRANSPORT_METRICS_NATS_URL=${CHORUS_TRANSPORT_METRICS_NATS_URL:-}
      - CHORUS_TRANSPORT_MODE=${CHORUS_TRANSPORT_MODE:-quic_only}
      - CHORUS_PROMPTS_DIR=/etc/chorus/prompts
      - CHORUS_DEFAULT_INSTRUCTIONS_PATH=/etc/chorus/prompts/defaults.md
      - CHORUS_ROLE=${CHORUS_ROLE:-arbiter}
    secrets:
      - chorus_license_id
      - resetdata_api_key
    configs:
      - source: chorus_bootstrap
        target: /config/bootstrap.json
    volumes:
      - chorus_data:/app/data
      - /rust/containers/WHOOSH/prompts:/etc/chorus/prompts:ro
      - /rust/containers/CHORUS/models.yaml:/app/configs/models.yaml:ro
    ports:
      - "${CHORUS_P2P_PORT:-9000}:9000/tcp"
      - "${CHORUS_P2P_PORT:-9000}:9000/udp"
    deploy:
      labels:
        - shepherd.autodeploy=true
      mode: replicated
      replicas: ${CHORUS_REPLICAS:-20}
      update_config:
        parallelism: 1
        delay: 10s
        failure_action: pause
        order: start-first
      restart_policy:
        condition: on-failure
        delay: 5s
        max_attempts: 3
        window: 120s
      resources:
        limits:
          cpus: "${CHORUS_CPU_LIMIT:-1.0}"
          memory: "${CHORUS_MEMORY_LIMIT:-4G}"
        reservations:
          cpus: "0.2"
          memory: 128M
      placement:
        preferences:
          - spread: node.hostname
    networks:
      - tengig
      - chorus_ipvlan
    extra_hosts:
      - "host.docker.internal:host-gateway"
    logging:
      driver: "json-file"
      options:
        max-size: "10m"
        max-file: "3"
        tag: "{{.ImageName}}/{{.Name}}/{{.ID}}"
    healthcheck:
      test: ["CMD", "curl", "-f", "http://localhost:8081/health"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 30s  # Increased from 10s to allow P2P mesh formation (15s bootstrap + margin)

  swoosh:
    image: anthonyrawlins/swoosh:1.0.2
    ports:
      - target: 8080
        published: 8800
        protocol: tcp
        mode: ingress
    environment:
      - SWOOSH_LISTEN_ADDR=:8080
      - SWOOSH_WAL_DIR=/data/wal
      - SWOOSH_SNAPSHOT_PATH=/data/snapshots/latest.json
    volumes:
      - swoosh_data:/data
    deploy:
      replicas: 1
      restart_policy:
        condition: on-failure
        delay: 5s
        max_attempts: 3
        window: 120s
      update_config:
        parallelism: 1
        delay: 10s
        failure_action: pause
        monitor: 60s
        order: start-first
      resources:
        limits:
          memory: 256M
          cpus: '0.5'
        reservations:
          memory: 128M
          cpus: '0.25'
      labels:
        - traefik.enable=true
        - traefik.docker.network=tengig
        - traefik.http.routers.swoosh.rule=Host(`swoosh.chorus.services`)
        - traefik.http.routers.swoosh.entrypoints=web,web-secured
        - traefik.http.routers.swoosh.tls=true
        - traefik.http.routers.swoosh.tls.certresolver=letsencryptresolver
        - traefik.http.services.swoosh.loadbalancer.server.port=8080
        - shepherd.autodeploy=true
        - traefik.http.services.swoosh.loadbalancer.passhostheader=true
    networks:
      - tengig
      - chorus_ipvlan
    healthcheck:
      test: ["CMD", "wget", "--no-verbose", "--tries=1", "-O", "/dev/null", "http://localhost:8080/health"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 40s

  postgres:
    image: postgres:15-alpine
    environment:
      - POSTGRES_DB=whoosh
      - POSTGRES_USER=whoosh
      - POSTGRES_PASSWORD_FILE=/run/secrets/whoosh_db_password
      - POSTGRES_INITDB_ARGS=--auth-host=scram-sha-256
    secrets:
      - whoosh_db_password
    volumes:
      - whoosh_postgres_data:/var/lib/postgresql/data
    deploy:
      replicas: 1
      restart_policy:
        condition: on-failure
        delay: 5s
        max_attempts: 3
        window: 120s
#      placement:
#        constraints:
#          - node.hostname == ironwood
      resources:
        limits:
          memory: 512M
          cpus: '1.0'
        reservations:
          memory: 256M
          cpus: '0.5'
    networks:
      - tengig
      - chorus_ipvlan
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -h localhost -p 5432 -U whoosh -d whoosh"]
      interval: 30s
      timeout: 10s
      retries: 5
      start_period: 40s

  redis:
    image: redis:7-alpine
    command: sh -c 'redis-server --requirepass "$$(cat /run/secrets/redis_password)" --appendonly yes'
    secrets:
      - redis_password
    volumes:
      - whoosh_redis_data:/data
    deploy:
      replicas: 1
      restart_policy:
        condition: on-failure
        delay: 5s
        max_attempts: 3
        window: 120s
      placement:
        preferences:
          - spread: node.hostname
      resources:
        limits:
          memory: 128M
          cpus: '0.25'
        reservations:
          memory: 64M
          cpus: '0.1'
    networks:
      - chorus_ipvlan
    healthcheck:
      test: ["CMD", "sh", "-c", "redis-cli --no-auth-warning -a $$(cat /run/secrets/redis_password) ping"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 30s

  prometheus:
    image: prom/prometheus:latest
    command:
      - '--config.file=/etc/prometheus/prometheus.yml'
      - '--storage.tsdb.path=/prometheus'
      - '--web.console.libraries=/usr/share/prometheus/console_libraries'
      - '--web.console.templates=/usr/share/prometheus/consoles'
    volumes:
      - /rust/containers/CHORUS/monitoring/prometheus/prometheus.yml:/etc/prometheus/prometheus.yml:ro
      - /rust/containers/CHORUS/monitoring/prometheus:/prometheus
      - /rust/containers/CHORUS/observability/prometheus/alerts:/etc/prometheus/alerts:ro
    ports:
      - "9099:9090"
    deploy:
      replicas: 1
      labels:
        - traefik.enable=true
        - traefik.http.routers.prometheus.rule=Host(`prometheus.chorus.services`)
        - traefik.http.routers.prometheus.entrypoints=web,web-secured
        - traefik.http.routers.prometheus.tls=true
        - traefik.http.routers.prometheus.tls.certresolver=letsencryptresolver
        - traefik.http.services.prometheus.loadbalancer.server.port=9090
        - shepherd.autodeploy=true
    networks:
      - chorus_ipvlan
      - tengig
    healthcheck:
      test: ["CMD", "wget", "--no-verbose", "--tries=1", "--spider", "http://localhost:9090/-/ready"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 10s

  grafana:
    image: grafana/grafana:latest
    user: "1000:1000"
    environment:
      - GF_SECURITY_ADMIN_PASSWORD=${GRAFANA_ADMIN_PASSWORD:-admin}
      - GF_SERVER_ROOT_URL=https://grafana.chorus.services
    volumes:
      - /rust/containers/CHORUS/monitoring/grafana:/var/lib/grafana
    ports:
      - "3300:3000"
    deploy:
      replicas: 1
      labels:
        - traefik.enable=true
        - traefik.http.routers.grafana.rule=Host(`grafana.chorus.services`)
        - traefik.http.routers.grafana.entrypoints=web,web-secured
        - traefik.http.routers.grafana.tls=true
        - traefik.http.routers.grafana.tls.certresolver=letsencryptresolver
        - traefik.http.services.grafana.loadbalancer.server.port=3000
        - shepherd.autodeploy=true
    networks:
      - chorus_ipvlan
      - tengig
    healthcheck:
      test: ["CMD", "wget", "--no-verbose", "--tries=1", "--spider", "http://localhost:3000/api/health"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 10s

  backbeat-pulse:
    image: docker.io/anthonyrawlins/backbeat-pulse:latest
    command: >
      ./pulse
      -cluster=chorus-production
      -admin-port=8080
      -raft-bind=0.0.0.0:9000
      -data-dir=/data
      -nats=nats://backbeat-nats:4222
      -tempo=2
      -bar-length=8
      -log-level=info
    expose:
      - "8080"
      - "9000"
    healthcheck:
      test: ["CMD", "nc", "-z", "localhost", "8080"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 60s
    deploy:
      replicas: 1
      restart_policy:
        condition: on-failure
        delay: 30s
        max_attempts: 5
        window: 120s
      update_config:
        parallelism: 1
        delay: 30s
        failure_action: pause
        monitor: 60s
        order: start-first
      placement:
        preferences:
          - spread: node.hostname
      resources:
        limits:
          memory: 256M
          cpus: '0.5'
        reservations:
          memory: 128M
          cpus: '0.25'
      labels:
        - traefik.enable=true
        - traefik.http.routers.backbeat-pulse.rule=Host(`backbeat-pulse.chorus.services`)
        - traefik.http.routers.backbeat-pulse.tls=true
        - traefik.http.routers.backbeat-pulse.tls.certresolver=letsencryptresolver
        - traefik.http.services.backbeat-pulse.loadbalancer.server.port=8080
    networks:
      - chorus_ipvlan
      - tengig
    logging:
      driver: "json-file"
      options:
        max-size: "10m"
        max-file: "3"
        tag: "backbeat-pulse/{{.Name}}/{{.ID}}"

  backbeat-reverb:
    image: docker.io/anthonyrawlins/backbeat-reverb:latest
    command: >
      ./reverb
      -cluster=chorus-production
      -nats=nats://backbeat-nats:4222
      -bar-length=8
      -log-level=info
    expose:
      - "8080"
    deploy:
      replicas: 2
      restart_policy:
        condition: on-failure
        delay: 10s
        max_attempts: 3
        window: 120s
      update_config:
        parallelism: 1
        delay: 15s
        failure_action: pause
        monitor: 45s
        order: start-first
      placement:
        preferences:
          - spread: node.hostname
      resources:
        limits:
          memory: 512M
          cpus: '1.0'
        reservations:
          memory: 256M
          cpus: '0.5'
      labels:
        - traefik.enable=true
        - traefik.http.routers.backbeat-reverb.rule=Host(`backbeat-reverb.chorus.services`)
        - traefik.http.routers.backbeat-reverb.tls=true
        - traefik.http.routers.backbeat-reverb.tls.certresolver=letsencryptresolver
        - traefik.http.services.backbeat-reverb.loadbalancer.server.port=8080
    networks:
      - chorus_ipvlan
      - tengig
    logging:
      driver: "json-file"
      options:
        max-size: "10m"
        max-file: "3"
        tag: "backbeat-reverb/{{.Name}}/{{.ID}}"

  backbeat-nats:
    image: nats:2.9-alpine
    command: ["--jetstream"]
    deploy:
      replicas: 1
      restart_policy:
        condition: on-failure
        delay: 10s
        max_attempts: 3
        window: 120s
      placement:
        preferences:
          - spread: node.hostname
      resources:
        limits:
          memory: 256M
          cpus: '0.5'
        reservations:
          memory: 128M
          cpus: '0.25'
    networks:
      - chorus_ipvlan
    logging:
      driver: "json-file"
      options:
        max-size: "10m"
        max-file: "3"
        tag: "nats/{{.Name}}/{{.ID}}"

  shepherd:
    image: containrrr/shepherd:latest
    environment:
      SLEEP_TIME: "5m"
      FILTER_SERVICES: "label=shepherd.autodeploy=true"
      WITH_REGISTRY_AUTH: "true"
      ROLLBACK_ON_FAILURE: "true"
      TZ: "UTC"
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
    deploy:
      replicas: 1
      restart_policy:
        condition: any
      placement:
        constraints:
          - node.role == manager

  hmmm-monitor:
    image: docker.io/anthonyrawlins/hmmm-monitor:latest
    environment:
      - WHOOSH_API_BASE_URL=http://swoosh:8080
    ports:
      - "9001:9001"
    deploy:
      labels:
        - shepherd.autodeploy=true
      replicas: 1
      restart_policy:
        condition: on-failure
        delay: 5s
        max_attempts: 3
        window: 120s
      resources:
        limits:
          memory: 128M
          cpus: '0.25'
        reservations:
          memory: 64M
          cpus: '0.1'
    networks:
      - chorus_ipvlan
    logging:
      driver: "json-file"
      options:
        max-size: "10m"
        max-file: "3"
        tag: "hmmm-monitor/{{.Name}}/{{.ID}}"

volumes:
  prometheus_data:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /rust/containers/CHORUS/monitoring/prometheus
  prometheus_config:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /rust/containers/CHORUS/monitoring/prometheus
  grafana_data:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /rust/containers/CHORUS/monitoring/grafana
  chorus_data:
    driver: local
  swoosh_data:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /rust/containers/SWOOSH/data
  whoosh_postgres_data:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /rust/containers/WHOOSH/postgres
  whoosh_redis_data:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /rust/containers/WHOOSH/redis
  whoosh_ui:
    driver: local
    driver_opts:
      type: none
      o: bind
      device: /rust/containers/WHOOSH/ui

networks:
  tengig:
    external: true

  chorus_ipvlan:
    external: true

configs:
  chorus_bootstrap:
    file: ./bootstrap.json

secrets:
  chorus_license_id:
    external: true
    name: chorus_license_id
  resetdata_api_key:
    external: true
    name: resetdata_api_key_v2
  whoosh_db_password:
    external: true
    name: whoosh_db_password
  gitea_token:
    external: true
    name: gitea_token
  webhook_token:
    external: true
    name: whoosh_webhook_token
  jwt_secret:
    external: true
    name: whoosh_jwt_secret_v4
  service_tokens:
    external: true
    name: whoosh_service_tokens
  redis_password:
    external: true
    name: whoosh_redis_password