docker-compose.yml

version: '3.5'

volumes:
  geonetwork:
  geonetwork_jetty:
  esdata:
  esbackup:
  kibanadata:
  pgdata:
  pglog:
  pgadmindata:
  cantaloupdata:
  portainer_data:
  previewdata:


services:
  traefik:
    image: traefik:v2.7
    ports:
      - 80:80
      - 443:443
      # Port 8080 is open by default on lal vm
      - 8080:8080
    volumes:
      # First volume needed for Traefik to listen to Docker events
      - /var/run/docker.sock:/var/run/docker.sock:ro
      - ./config/traefik/ssl:/etc/certs:ro
      - ./config/traefik/traefik.yml:/etc/traefik/traefik.yml:ro
      - ./config/traefik/traefik-config.yml:/etc/traefik/config.yml:ro
    restart: always
    labels:
      - traefik.enable=true
      - traefik.http.routers.traefik.entryPoints=websecure
      - traefik.http.routers.traefik.tls=true
      - traefik.http.routers.traefik.service=api@internal
      - traefik.http.routers.traefik.rule=Host(`traefik.geohistoricaldata.org`, `www.traefik.geohistoricaldata.org`)


  portainer:
    image: portainer/portainer-ce:latest
    command: -H unix:///var/run/docker.sock
    restart: always
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
      - portainer_data:/data
    labels:
      - traefik.enable=true
      - traefik.http.routers.frontend.rule=Host(`portainer.geohistoricaldata.org`, `www.portainer.geohistoricaldata.org`)
      - traefik.http.routers.frontend.entrypoints=websecure
      - traefik.http.services.frontend.loadbalancer.server.port=9000
      - traefik.http.routers.frontend.service=frontend
      - traefik.http.routers.frontend.tls=true


  database:
    image: postgresql-pgcron
    restart: always
    environment:
      POSTGRES_USER: geonetwork
      POSTGRES_PASSWORD: $POSTGRES_PASSWORD
      POSTGRES_DB: geonetwork
    command: [ "postgres",
               "-c", "log_statement=all",
               "-c", "logging_collector=true",
               "-c", "log_file_mode=0644",
               "-c", "log_directory=/var/log/postgresql",
               "-c", "log_filename=postgresql.log",
               "-c", "shared_preload_libraries=pg_cron",
               "-c", "cron.database_name=geonetwork",
      	       "-c", "cron.log_run=off",
               "-c", "cron.timezone=Europe/Paris", ]
    volumes:
      - pgdata:/var/lib/postgresql/data
      - pglog:/var/log/postgresql
    labels:
      - traefik.enable=true
      - traefik.http.routers.database.entryPoints=websecure
      - traefik.tcp.routers.database.rule=HostSNI(`*`,`database.geohistoricaldata.org`)
      - traefik.tcp.routers.database.entryPoints=database
      - traefik.tcp.routers.database.service=database
      - traefik.tcp.services.database.loadbalancer.server.port=5432
      - traefik.http.routers.database.tls=true


  pgadmin4:
    image: dpage/pgadmin4:7.8
    restart: always
    environment:
      PGADMIN_DEFAULT_EMAIL: $PGADMIN_DEFAULT_EMAIL
      PGADMIN_DEFAULT_PASSWORD: $PGADMIN_DEFAULT_PASSWORD
      # SCRIPT_NAME needed for url
      SCRIPT_NAME: /pgadmin
    # ports need to stay, even with traefik, or it will cause a security error
    ports:
      - 5050:80
    volumes:
      - pgadmindata:/var/lib/pgadmin
    labels:
      - traefik.enable=true
      - traefik.http.routers.pgadmin4.entryPoints=websecure
      - traefik.http.routers.pgadmin4.rule=Host(`pgadmin.geohistoricaldata.org`, `www.pgadmin.geohistoricaldata.org`)
      - traefik.http.routers.pgadmin4.tls=true
      - traefik.http.routers.pgadmin4.middlewares=redirect-pgadmin
      # regex trouvé ici : https://gist.github.com/kekru/d088be6a3fa844089ae62d80c077bb38
      - traefik.http.middlewares.redirect-pgadmin.redirectregex.regex=^https:\/\/([^\/]+)\/?$$
      - traefik.http.middlewares.redirect-pgadmin.redirectregex.replacement=https://$$1/pgadmin


  geonetwork:
    image: geonetwork:4.2.5
    restart: always
    environment:
      DATA_DIR: /catalogue-data

      JAVA_OPTS: >
        -Dorg.eclipse.jetty.annotations.AnnotationParser.LEVEL=OFF
        -Djava.security.egd=file:/dev/./urandom -Djava.awt.headless=true
        -Xms512M -Xss512M -Xmx2G -XX:+UseConcMarkSweepGC
        -Dgeonetwork.resources.dir=/catalogue-data/resources
        -Dgeonetwork.data.dir=/catalogue-data
        -Dgeonetwork.codeList.dir=/var/lib/jetty/webapps/geonetwork/WEB-INF/data/config/codelist
        -Dgeonetwork.schema.dir=/var/lib/jetty/webapps/geonetwork/WEB-INF/data/config/schema_plugins

      ES_HOST: elasticsearch
      ES_PROTOCOL: http
      ES_PORT: 9200

      KB_URL: http://kibana:5601

      GEONETWORK_DB_TYPE: postgres
      GEONETWORK_DB_HOST: database
      GEONETWORK_DB_PORT: 5432
      GEONETWORK_DB_NAME: geonetwork
      GEONETWORK_DB_USERNAME: geonetwork
      GEONETWORK_DB_PASSWORD: $POSTGRES_PASSWORD
    volumes:
      - geonetwork:/catalogue-data
      - geonetwork_jetty:/var/lib/jetty
      - ./config/geonetwork/config-security-core.xml:/var/lib/jetty/webapps/geonetwork/WEB-INF/config-security/config-security-core.xml
    depends_on:
      - elasticsearch
      - database
    labels:
      - traefik.enable=true
      - traefik.http.routers.geonetwork.entryPoints=websecure
      - traefik.http.routers.geonetwork.rule=Host(`catalog.geohistoricaldata.org`, `www.catalog.geohistoricaldata.org`)
      - traefik.http.routers.geonetwork.tls=true
      - traefik.http.routers.geonetwork.middlewares=redirect-geonetwork
      - traefik.http.middlewares.redirect-geonetwork.redirectregex.regex=^https:\/\/([^\/]+)\/?$$
      - traefik.http.middlewares.redirect-geonetwork.redirectregex.replacement=https://$$1/geonetwork


  elasticsearch:
    image: elasticsearch:7.17.13
    restart: always
    ulimits:
      memlock:
        soft: -1
        hard: -1
      nofile:
        soft: 65536
        hard: 65536
    environment:
      ES_JAVA_OPTS: "-Xms1G -Xmx1G"
      discovery.type: single-node
      path.repo: "/usr/share/elasticsearch/snapshots"
    volumes:
      - esdata:/usr/share/elasticsearch/data
      - esbackup:/usr/share/elasticsearch/snapshots


  kibana:
    image: kibana:7.9.2
    restart: always
    environment:
      SERVER_NAME: 'kibana'
      ELASTICSEARCH_URL: http://elasticsearch:9200/
      SERVER_BASEPATH: /geonetwork/dashboards
      SERVER_REWRITEBASEPATH: 'false'
      KIBANA_INDEX: .dashboards
      XPACK_MONITORING_UI_CONTAINER_ELASTICSEARCH_ENABLED: 'true'
    volumes:
      - kibanadata:/usr/share/kibana/data


  preview:
    image: nginx
    volumes:
      - /data/soduco_webmap-main:/usr/share/nginx/html:ro
      - ./config/directory/default.conf:/etc/nginx/conf.d/default.conf
    restart: unless-stopped
    labels:
      - traefik.enable=true
      - traefik.http.routers.preview.entryPoints=websecure
      - traefik.http.routers.preview.rule=Host(`preview.geohistoricaldata.org`, `www.preview.geohistoricaldata.org`)
      - traefik.http.routers.preview.tls=true


  ontop:
    image: ontop/ontop:5.1.1
    restart: unless-stopped
    environment:
      ONTOP_ONTOLOGY_FILE: /opt/ontop/input/catalog.ttl
      ONTOP_MAPPING_FILE: /opt/ontop/input/catalog.obda
      ONTOP_PROPERTIES_FILE: /opt/ontop/input/catalog.properties
      # ONTOP_PORTAL_FILE: /opt/ontop/input/university-complete.portal.toml
      ONTOP_CORS_ALLOWED_ORIGINS: "*"
      ONTOP_DEV_MODE: "true"
      ONTOP_LAZY_INIT: "true"
    volumes:
      # files needed by ontop
      - ./config/ontop/input:/opt/ontop/input
      - ./config/ontop/jdbc:/opt/ontop/jdbc
    depends_on:
      - "database"
    labels:
      - traefik.enable=true
      - traefik.http.routers.ontop.entryPoints=websecure
      - traefik.http.routers.ontop.rule=Host(`rdf.geohistoricaldata.org`, `www.rdf.geohistoricaldata.org`)
      - traefik.http.routers.ontop.tls=true
      - traefik.http.routers.ontop.middlewares=redirect-lodview
      # middleware to redirect resources to lodview.
      # "(?:/lodview)*" prevent an infinite loop of /lodview
      # (/.*/.{8}-.{4}-.{4}-.{4}-.{12}) match with /something/geonetwork_uuid
      # example : https://rdf.geohistoricaldata.org/lodview/instantiation/e5a2978d-8b05-401d-ad0c-5b91d1f273d3.html
      - traefik.http.middlewares.redirect-lodview.redirectregex.regex=^https://rdf.geohistoricaldata.org(?:/lodview)*(/.*/.{8}-.{4}-.{4}-.{4}-.{12})
      - traefik.http.middlewares.redirect-lodview.redirectregex.replacement=https://rdf.geohistoricaldata.org/lodview$${1}
      - traefik.http.routers.ontop-sparql.entryPoints=websecure
      # Entrypoint for sparql endpoint. Necessary for lodview redirection rule to work properly
      - traefik.http.routers.ontop-sparql.rule=Host(`rdf.geohistoricaldata.org`, `www.rdf.geohistoricaldata.org`) && PathPrefix(`/sparql`)
      - traefik.http.routers.ontop-sparql.tls=true
      # Entrypoint and middleware to rewrite Ontop responses into LodLive expected format. (Ontop's sparql endpoint doesn't take callbacks in http request)
      - traefik.http.routers.ontop-rewrite.rule=Host(`rdf.geohistoricaldata.org`, `www.rdf.geohistoricaldata.org`) && Query(`callback=lodlive`)
      - traefik.http.routers.ontop-rewrite.entryPoints=websecure
      - traefik.http.routers.ontop-rewrite.tls=true
      - traefik.http.routers.ontop-rewrite.middlewares=rewrite-lodlive
      - traefik.http.middlewares.rewrite-lodlive.plugin.rewrite.lastModified=true
      - traefik.http.middlewares.rewrite-lodlive.plugin.rewrite.monitor.types[0]=text/html
      - traefik.http.middlewares.rewrite-lodlive.plugin.rewrite.rewrites[0].regex=(?s:(.*))
      - traefik.http.middlewares.rewrite-lodlive.plugin.rewrite.rewrites[0].replacement=lodlive($${1})


  lodview:
    image: lodview:latest
    restart: unless-stopped
    labels:
      - traefik.enable=true
      - traefik.http.routers.lodview.entryPoints=websecure
      - traefik.http.routers.lodview.rule=Host(`rdf.geohistoricaldata.org`, `www.rdf.geohistoricaldata.org`) && PathPrefix(`/lodview`)
      - traefik.http.routers.lodview.tls=true

# Ontop for Solenn's work.
# Container's name must be changed to something better, like ontop_dir
  ontop_rdf:
    image: ontop/ontop-endpoint:4.2.0
    restart: unless-stopped
    environment:
      ONTOP_ONTOLOGY_FILE: /opt/ontop/input/directory.rdf
      ONTOP_MAPPING_FILE: /opt/ontop/input/directory.ttl
      ONTOP_PROPERTIES_FILE: /opt/ontop/input/directory.properties
      # ONTOP_PORTAL_FILE: /opt/ontop/input/university-complete.portal.toml
      ONTOP_CORS_ALLOWED_ORIGINS: "*"
      ONTOP_DEV_MODE: "true"
      ONTOP_LAZY_INIT: "true"
    volumes:
      # files needed by ontop
      - ./config/ontop_photographes/input:/opt/ontop/input
      - ./config/ontop_photographes/jdbc:/opt/ontop/jdbc
    labels:
      - traefik.enable=true
      - traefik.http.routers.ontop_rdf.entryPoints=websecure
      - traefik.http.routers.ontop_rdf.rule=Host(`dir.geohistoricaldata.org`, `www.dir.geohistoricaldata.org`)
      - traefik.http.routers.ontop_rdf.tls=true


  cantaloupe:
    image: uclalibrary/cantaloupe:5.0.5-7
    restart: unless-stopped
    cpus: 4
    mem_limit: 16g
    environment:
      CANTALOUPE_ENDPOINT_ADMIN_ENABLED: "true"
      CANTALOUPE_ENDPOINT_ADMIN_SECRET: $CANTALOUPE_PASSWORD
    volumes:
      - /data/cantaloupe:/imageroot
      - ./config/cantaloupe/cantaloupe.properties:/etc/cantaloupe.properties
    labels:
      - traefik.enable=true
      - traefik.http.routers.cantaloupe.entryPoints=websecure
      - traefik.http.routers.cantaloupe.rule=Host(`iiif.geohistoricaldata.org`, `www.iiif.geohistoricaldata.org`)
      - traefik.http.routers.cantaloupe.tls=true
      - traefik.http.services.cantaloupe.loadbalancer.server.port=8182

  geonetwork-ui:
    image: geonetwork/geonetwork-ui-datahub:2.0.1
    restart: unless-stopped
    environment:
      - GN4_API_URL=https://catalog.geohistoricaldata.org/geonetwork/srv/api
    labels:
      - traefik.enable=true
      - traefik.http.routers.geonetwork-ui.entryPoints=websecure
      - traefik.http.routers.geonetwork-ui.rule=Host(`catalog.geohistoricaldata.org`, `www.catalog.geohistoricaldata.org`) && PathPrefix(`/datahub`)
      - traefik.http.routers.geonetwork-ui.tls=true

  map-proxy:
    image: ghcr.io/mapproxy/mapproxy/mapproxy:1.16.0-nginx
    restart: unless-stopped
    volumes:
      - /data/config/map_proxy/mapproxy.yaml:/mapproxy/config/mapproxy.yaml:ro
      - /data/config/map_proxy/cache_data:/mapproxy/config/cache_data
    labels:
      - traefik.enable=true
      - traefik.http.routers.map-proxy.entryPoints=websecure
      - traefik.http.routers.map-proxy.rule=Host(`map.geohistoricaldata.org`, `www.map.geohistoricaldata.org`)
      - traefik.http.routers.map-proxy.tls=true

  directory:
    image: nginx
    volumes:
      - /data/directory:/usr/share/nginx/html:ro
      - ./config/directory/default.conf:/etc/nginx/conf.d/default.conf
    labels:
      - traefik.enable=true
      - traefik.http.routers.directory.entryPoints=websecure
      - traefik.http.routers.directory.rule=Host(`directory.geohistoricaldata.org`, `www.directory.geohistoricaldata.org`)
      - traefik.http.routers.directory.tls=true