Glances does not export data to prometheus with docker-compose

[A7F]

Describe the bug
I am setting up a stack with glances, prometheus and grafana using docker compose. I started with an example provided here which did not work as I got a 401 unauthorized error on Influxdb, which is why I decided to try Prometheus instead.
So this is the docker file I am currently working with

services:
  glances:
    container_name: glances
    image: nicolargo/glances:dev
    restart: unless-stopped
    environment:
      - GLANCES_OPT=-q --export prometheus --time 10
    pid: "host"
    privileged: true
    network_mode: "host"
    depends_on:
      - prometheus
    ports:
      - 9091:9091
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock:ro
      - ./cfg/glances/glances.conf:/glances/conf/glances.conf
     

  prometheus:
    image: prom/prometheus:latest
    restart: unless-stopped
    container_name: prometheus
    ports:
    - "9090:9090"
    volumes:
    - ./cfg/prometheus/prometheus.yml:/etc/prometheus/prometheus.yml
    - prometheus_data:/pormetheus

  grafana:
    container_name: grafana
    image: grafana/grafana
    restart: unless-stopped
    ports:
      - 8080:3000
    user: "1000"
    depends_on:
      - prometheus
    volumes:
      - grafana_data:/var/lib/grafana
    networks:
      - grafana_network

volumes:
  # define persistence volumes
  prometheus_data:
  grafana_data:

networks:
  grafana_network:
    name: grafana_network
    driver: bridge
    driver_opts:
      com.docker.network.bridge.name: br_grafana

and the configuration files are

global:
 scrape_interval: 15s

scrape_configs:
  - job_name: 'glances_exporter'
    scrape_interval: 5s
    static_configs:
      - targets: ['0.0.0.0:9091']

and glances specific section

[prometheus]
host=0.0.0.0
port=9091
#prefix=glances
labels=src:glances

Expected behavior
I expected to see at least something in prometheus time series database, but I get nothing. What is interesting is that, in glances container, I get

2025-07-10 08:29:37,532 -- INFO -- Start Glances 4.4.0_dev1
2025-07-10 08:29:37,533 -- INFO -- CPython 3.12.11 (/venv/bin/python3) and psutil 7.0.0 detected
2025-07-10 08:29:37,551 -- INFO -- Read configuration file '/etc/glances/glances.conf'
2025-07-10 08:29:38,846 -- INFO -- Start GlancesStandalone mode
2025-07-10 08:29:40,009 -- INFO -- Start Prometheus exporter on localhost:9091
2025-07-10 08:29:40,908 -- INFO -- Quiet mode is ON, nothing will be displayed

what seems odd to me is that it keeps saying "start prometheus exporter on localhost", even though I said 0.0.0.0 in configuration file AND, if I change it to some other ip address in config file, stays fixed to localhost. I tried passing the actual host ip address (changing it in prometheus as well) but still no luck. Am I missing something?

Environement

• Operating System (lsb_release -a or OS name/version): Dockerized on Windows 11
• Glances & psutil versions: glances 4.4.0_dev1, psutil 7.0.0
• How do you install Glances (Pypi package, script, package manager, source): docker compose

[nicolargo]

Hi @A7F

Please can you run the command in debug mode and copy paste the result of the log file ?

[A7F]

yes of course, I copy pasted it on Gist at this link. If I could add to the issue, looks like the same thing happens if I try to export to mongodb, as it is unable to authenticate. Weird thing is, in all cases tries to connect to localhost, regardless of the address I'm passing in glances config file...

[nicolargo]

It look like the Glances.conf file is not taken into account.

Try this:

GLANCES_OPT=-q --export prometheus --time 10 -C /glances/conf/glances.conf

[A7F]

it works; still having issues with exporting the data (in glances I am posting to 0.0.0.0:9091, prometheus is scraping from localhost:9091) but the fact that glances is picking up the correct config is already an awesome thing.
Still the need for specifying the config file location puzzles me quite a bit as I am mounting the volume properly, but I'm okay with that!