Testing a role which uses elasticsearch (reload systemd)

[lksnmnn]

Hi there,

I am working on covering all our playbooks using molecule in YOLO mode (outside of roles). For most of our playbooks/roles it works very nicely and I will share my setup here once it works.

BUT I am stuck when testing a playbook which installs elastic search using the official ansible role. Aparently, there is an issue when the role attempts to reload the systemd daemon after installing ES:

    RUNNING HANDLER [elastic.elasticsearch : reload systemd configuration] *********
    task path: /root/.cache/molecule/infra/default/roles/elastic.elasticsearch/handlers/main.yml:2
fatal: [test_host]: FAILED! => {
    "changed": false,
    "invocation": {
        "module_args": {
            "daemon_reexec": false,
            "daemon_reload": true,
            "enabled": null,
            "force": null,
            "masked": null,
            "name": null,
            "no_block": false,
            "scope": null,
            "state": null,
            "user": null
        }
    },
    "msg": "failure 1 during daemon-reload: Failed to connect to bus: No such file or directory\n"
}

Here parts of my current molecule.yaml (I tested different settings, but it wont work):

driver:
  name: docker

platforms:
  - name: test_host
    image: geerlingguy/docker-ubuntu1804-ansible:latest
    pre_build_image: true
    privileged: true
    tty: true
    volume_mounts:
      - "/sys/fs/cgroup:/sys/fs/cgroup:ro"
      - /run
      - /tmp
    command: '/sbin/init'
    environment:
      container: docker
    groups:
      - group1
      - group2

Note: I am running Molecule on Mac using the offical docker image.

Invoke:

 docker run --rm -it \   
    -v "$(pwd)":/tmp/$(basename "${PWD}"):ro \
    -v /var/run/docker.sock:/var/run/docker.sock \
    quay.io/ansible/molecule:3.0.6 \
    molecule test

Thank you in advance.

Cheers,
Lukas

[lksnmnn]

FYI, if I refrain from overwriting the containers command, I get the error:

msg": "failure 1 during daemon-reload: System has not been booted with systemd as init system (PID 1). Can't operate.

[bbaassssiiee]

Quote: "Works on my cgroups!" Try adding cgroupns_mode: host, and the volume /var/lib/containerd

Snippet from my molecule.yml:

dependency:
  name: galaxy
driver:
  name: docker
platforms:
  - name: my-server
    image: almalinux:8
    command: /sbin/init
    capabilities:
      - SYS_ADMIN
    cgroupns_mode: host
    tmpfs:
      - /run
      - /tmp
    volumes:
      - /sys/fs/cgroup:/sys/fs/cgroup:rw
      - /var/lib/containerd