metalfrom.eu/docker-compose.dev.yml
Nicolas Fryder 72a52a8d3b feat(geocoder): seuil de confiance strict, dedup, purge ancien pipeline
Durcissement fiabilité (suite audit).

Confiance & granularité
- worker: un résultat n'est accepté ('done') que si confiance >= GEOCODE_MIN_CONFIDENCE
  (0.7) ET granularité non-grossière (rejette country/state/county/region). Sinon
  on continue les fallbacks, puis -> llm_needed.
- fix majeur: sur cache hit la confiance était écrite 0.5 en dur (97% des lignes
  faussées). Elle est désormais lue depuis geocode_cache (recalculée du raw).
- migration 012: colonnes confidence+granularity sur geocode_cache (recalcul des
  entrées Geoapify depuis le raw), geocode_granularity sur band_locations.

Dedup (7x moins de travail)
- fast-path: un band_location dont le (lieu,pays) est déjà 'done' copie le
  résultat sans appel API. Index fonctionnel lower(location_raw).

Ancien pipeline retiré
- geocode_queue n'est plus lu nulle part (endpoints /geocoding/reset-errors et
  /requeue-all supprimés, /live et /geocoding et Monitor basculés sur
  band_locations, panneau admin "ancien pipeline" retiré).

Boutons reset (onglet Géocodage, zone dangereuse)
- POST /admin/api/locations/reset-all: remet tout en queue + efface coords
- POST /admin/api/geocode-cache/purge-nominatim: purge le cache Nominatim

Divers
- groq_worker: coût calculé par modèle (70B vs 8B) au lieu du tarif 70B fixe
- GEOCODE_MIN_CONFIDENCE ajouté aux deux compose

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-02 20:32:16 +02:00

170 lines
4.6 KiB
YAML

services:
redis:
image: redis:7-alpine
command: ["redis-server", "--appendonly", "yes"]
volumes:
- bm_dev_redis_data:/data
networks:
- coolify
flaresolverr:
image: ghcr.io/flaresolverr/flaresolverr:latest
environment:
LOG_LEVEL: info
CAPTCHA_SOLVER: none
networks:
- coolify
labels:
- traefik.enable=false
restart: unless-stopped
crawler:
build:
context: apps/crawler
environment:
DATABASE_URL: ${DATABASE_URL}
FLARESOLVERR_URL: http://flaresolverr:8191
CRAWLER_SCHED_INCREMENTAL_H: "4"
CRAWLER_SCHED_ENRICH_H: "2"
CRAWLER_SCHED_FULL_DAY: "0"
depends_on:
- flaresolverr
networks:
- coolify
labels:
- traefik.enable=false
restart: unless-stopped
geocoder-enqueue:
build:
context: apps/geocoder
working_dir: /app
environment:
DATABASE_URL: ${DATABASE_URL}
GEOCODE_ENQUEUE_BATCH: "100000"
GEOCODE_ENQUEUE_POLL: "15"
command: ["python", "src/enqueue.py"]
networks:
- coolify
labels:
- traefik.enable=false
restart: unless-stopped
geocoder-worker:
build:
context: apps/geocoder
working_dir: /app
environment:
DATABASE_URL: ${DATABASE_URL}
GEOAPIFY_API_KEY: ${GEOAPIFY_API_KEY}
GEOCODER_MIN_DELAY: "0.22"
GEOCODER_JITTER: "0.10"
GEOCODE_MAX_PER_RUN: "100000"
GEOCODE_MAX_GEO_TRIES: "3"
GEOCODE_MIN_CONFIDENCE: "0.7"
command: ["python", "src/worker.py"]
networks:
- coolify
labels:
- traefik.enable=false
restart: unless-stopped
groq-worker:
build:
context: apps/geocoder
working_dir: /app
environment:
DATABASE_URL: ${DATABASE_URL}
GROQ_API_KEY: ${GROQ_API_KEY}
GROQ_MAX_TRIES: "5"
GROQ_CALL_DELAY: "2.5"
command: ["python", "src/groq_worker.py"]
networks:
- coolify
labels:
- traefik.enable=false
restart: unless-stopped
pgadmin:
image: dpage/pgadmin4:8
environment:
PGADMIN_DEFAULT_EMAIL: ${PGADMIN_EMAIL}
PGADMIN_DEFAULT_PASSWORD: ${PGADMIN_PASSWORD}
PGADMIN_CONFIG_SERVER_MODE: "True"
volumes:
- bm_dev_pgadmin_data:/var/lib/pgadmin
networks:
- coolify
labels:
- traefik.enable=true
- traefik.docker.network=coolify
- traefik.http.routers.dev-pgadmin.rule=Host(`dev-pgadmin.metalfrom.eu`)
- traefik.http.routers.dev-pgadmin.entrypoints=https
- traefik.http.routers.dev-pgadmin.tls=true
- traefik.http.routers.dev-pgadmin.tls.certresolver=letsencrypt
- traefik.http.services.dev-pgadmin.loadbalancer.server.port=80
api:
build:
context: apps/api
environment:
PORT: "3000"
DATABASE_URL: ${DATABASE_URL}
BM_IMPORT_TOKEN: ${BM_IMPORT_TOKEN}
CORS_ORIGINS: ${CORS_ORIGINS:-https://dev.metalfrom.eu}
ADMIN_JWT_SECRET: ${ADMIN_JWT_SECRET}
ADMIN_SEED_USERNAME: ${ADMIN_SEED_USERNAME}
ADMIN_SEED_PASSWORD_HASH: ${ADMIN_SEED_PASSWORD_HASH}
networks:
coolify:
aliases:
- dev-api-internal
labels:
- traefik.enable=true
- traefik.docker.network=coolify
- traefik.http.routers.dev-api.rule=Host(`dev-api.metalfrom.eu`)
- traefik.http.routers.dev-api.entrypoints=https
- traefik.http.routers.dev-api.tls=true
- traefik.http.routers.dev-api.tls.certresolver=letsencrypt
- traefik.http.services.dev-api.loadbalancer.server.port=3000
admin:
build:
context: apps/admin
args:
API_UPSTREAM: http://dev-api-internal:3000
networks:
- coolify
labels:
- traefik.enable=true
- traefik.docker.network=coolify
- traefik.http.routers.dev-admin.rule=Host(`admin.dev.metalfrom.eu`)
- traefik.http.routers.dev-admin.entrypoints=https
- traefik.http.routers.dev-admin.tls=true
- traefik.http.routers.dev-admin.tls.certresolver=letsencrypt
- traefik.http.services.dev-admin.loadbalancer.server.port=80
web:
build:
context: apps/web
args:
API_BASE: https://dev-api.metalfrom.eu
networks:
- coolify
labels:
- traefik.enable=true
- traefik.docker.network=coolify
- traefik.http.routers.dev-web.rule=Host(`dev.metalfrom.eu`)
- traefik.http.routers.dev-web.entrypoints=https
- traefik.http.routers.dev-web.tls=true
- traefik.http.routers.dev-web.tls.certresolver=letsencrypt
- traefik.http.services.dev-web.loadbalancer.server.port=80
networks:
coolify:
external: true
name: coolify
volumes:
bm_dev_redis_data:
bm_dev_pgadmin_data: