Can't reach out DB deployed via Coolify from within Compose stack.

[Stf-F]

Hello,

EDIT: As I have kept testing, I have realised that the Supabase image keeps reverting to listening to localhost/127.0.0.1 when SSL full verify is on. It listens to 0.0.0.0 with no SSL. I have tested the behaviour of the standard PG 17 image and there the container always listens 0.0.0.0, with or without SSL. Furthermore the Supabase DB was spitting out a few errors:

host: 'hs48k0g8kswkwwkcgg0wsock', port: 5432, hostaddr: '10.0.1.2': connection failed: connection to server at "10.0.1.2", port 5432 failed: root certificate file "/etc/ssl/certs/coolify-ca.crt" does not exist
Either provide the file or change sslmode to disable server certificate verification.

So looks like something wrong/odd with the Supabase image, and was there or thereabout a while ago without realising it.

---

Apologies if this has been already answered somewhere else, I genuinely couldn't find an answer to it anywhere. I have been at it for about a day and still can't figure out what's happening.

I have a compose setup inherited from CookieCutter and use to deploy a Django app that I want to connect to a DB deployed via Coolify, and not via Compose. (Supabase image)

According to this doc: https://coolify.io/docs/knowledge-base/docker/compose#connect-to-predefined-networks

I need:

• to use the full name of the DB as it appears in the UI as a host in my connection string, so in my case postgresql-database-hs48k0g8kswkwwkcgg0wsock
• switch on the Connect To Predefined Network option in the Django stack. Done.

Using docker network inspect coolify I have verified that my DB and my app containers are on the same network, and they are.
Although I see that the "Name" of my DB in the output is hs48k0g8kswkwwkcgg0wsock and not postgresql-database-hs48k0g8kswkwwkcgg0wsock

Since I created a custom user and custom db I am using a connection string like this in my DATABASE_URL env var:

postgres://CUSTOM_USER:CUSTOM_PWD@postgresql-database-hs48k0g8kswkwwkcgg0wsock:5432/CUSTOM_DB?sslmode=verify-full&sslrootcert=/etc/ssl/certs/coolify-ca.crt

I have also tried with the Network log output Name, and also of course the standard string as it appears in the UI:

postgres://CUSTOM_USER:CUSTOM_PWD@hs48k0g8kswkwwkcgg0wsock:5432/CUSTOM_DB?sslmode=verify-full&sslrootcert=/etc/ssl/certs/coolify-ca.crt

But no dice, I keep getting signs that the Django container can't talk to my Postgres.

host: 'hs48k0g8kswkwwkcgg0wsock', port: 5432, hostaddr: '10.0.1.8': connection failed: connection to server at "10.0.1.8", port 5432 failed: Connection refused

Now, it looks like Postgres is only listening on localhost, which somehow could explain why outsiders can't reach out.

2026-02-12 20:53:07.803 UTC [1] LOG: starting PostgreSQL 17.4 on aarch64-unknown-linux-gnu, compiled by gcc (GCC) 13.2.0, 64-bit
2026-02-12 20:53:07.810 UTC [1] LOG: listening on IPv6 address "::1", port 5432
2026-02-12 20:53:07.810 UTC [1] LOG: listening on IPv4 address "127.0.0.1", port 5432
2026-02-12 20:53:07.825 UTC [1] LOG: listening on Unix socket "/run/postgresql/.s.PGSQL.5432"

The strange thing is that even when making the DB public, I keep getting:

host: 'hs48k0g8kswkwwkcgg0wsock', port: 5432, hostaddr: 'fde4:d481:aa6e::8': connection failed: connection to server at "fde4:d481:aa6e::8", port 5432 failed: Connection refused
Is the server running on that host and accepting TCP/IP connections?

And the proxy logs indicate this:

2026/02/13 13:34:18 [error] 30#30: *21 no live upstreams while connecting to upstream, client: 10.0.1.1, server: 0.0.0.0:5432, upstream: "hs48k0g8kswkwwkcgg0wsock", bytes from/to client:0/0, bytes from/to upstream:0/0

Could anyone shed some light on this please? Because this leaves me completely puzzled.

Thanks 🙏

My compose for reference with the bits taken out from CookieCutter.

volumes:
  production_postgres_data: {}
  production_postgres_data_backups: {}
  # production_traefik: {} # commented as not using traefik

services:
  django:
    build:
      context: .
      dockerfile: ./compose/production/django/Dockerfile
    image: seo_optimiser_production_django
    # depends_on:
    # - postgres
    # - redis
    env_file:
      - ./.envs/.production/.django
      - ./.envs/.production/.postgres
    command: /start
    expose:
      - "5000" # <== I have added this. It is not part of the original setup

  awscli:
    build:
      context: .
      dockerfile: ./compose/production/aws/Dockerfile
    env_file:
      - ./.envs/.production/.django
    volumes:
      - production_postgres_data_backups:/backups:z

  qcluster:
    build:
      context: .
      dockerfile: ./compose/production/django/Dockerfile
    image: seo_optimiser_app:latest
    command: python manage.py qcluster
    env_file:
      - ./.envs/.production/.django
      - ./.envs/.production/.postgres
    # depends_on:
    #   - postgres
    restart: always
    depends_on:
      - django

  # postgres:
  #   build:
  #     context: .
  #     dockerfile: ./compose/production/postgres/Dockerfile
  #   image: seo_optimiser_production_postgres
  #   volumes:
  #     - production_postgres_data:/var/lib/postgresql/18/docker
  #     - production_postgres_data_backups:/backups
  #   env_file:
  #     - ./.envs/.production/.postgres

  # traefik:
  #   build:
  #     context: .
  #     dockerfile: ./compose/production/traefik/Dockerfile
  #   image: seo_optimiser_production_traefik
  #   depends_on:
  #     - django
  #   volumes:
  #     - production_traefik:/etc/traefik/acme
  #   ports:
  #     - "0.0.0.0:80:80"
  #     - "0.0.0.0:443:443"

  # redis:
  #   image: docker.io/redis:7.2