Hetzner’s cx11 server size not available in US locations

[quinncomendant]

Describe the bug

Algo fails on Hetzner’s USA locations when using thecx11 server size, which is the default included in Algo’s config.cfg file. The Hetzner Cloud API responds with:

error during placement (resource_unavailable)

This is not necessarily a bug in Algo, since the same error occurs when trying to create a cx11 server using Hetzner's CLI tools.

I asked Hetzner why this error occurs, and they replied:

CX-Servers with INTEL CPU are not offered in our US-Locations.

So, the fix is to update Algo's config.cfg file to use server cpx11 instead of cx11 (cpx11 is the smallest server size that is available in all Hetzner locations; it has 2 vCPUs instead of 1, and costs €3.85 instead of €3.29 per month):

  hetzner:
    server_type: cpx11
    image: ubuntu-20.04

To Reproduce

Steps to reproduce the behavior:

1. Run ./algo
2. Select Provider = Hetzner Cloud.
3. Select a region of ash or hil.
4. Ansible fails with the error: fatal: [localhost]: FAILED! => {"changed": false, "msg": "error during placement"}.

Expected behavior

Algo's default config should work for all providers and all locations.

Additional context

Here's the full ANSIBLE_VERBOSITY=4 error output to help people find this ticket:

TASK [cloud-hetzner : Create a server...] *******************************************************************************************************************************************
task path: /Users/quinn/src/algo/roles/cloud-hetzner/tasks/main.yml:16
The full traceback is:
  File "/var/folders/sd/sqyq8bqd6bj1vjh9v7f_5ddc0000gp/T/ansible_hcloud_server_payload_gf68n8k7/ansible_hcloud_server_payload.zip/ansible_collections/hetzner/hcloud/plugins/modules/hcloud_server.py", line 438, in _create_server
  File "/Users/quinn/src/algo/.venv/lib/python3.10/site-packages/hcloud/servers/client.py", line 583, in create
    response = self._client.request(url="/servers", method="POST", json=data)
  File "/Users/quinn/src/algo/.venv/lib/python3.10/site-packages/hcloud/hcloud.py", line 242, in request
    self._raise_exception_from_json_content(json_content)
  File "/Users/quinn/src/algo/.venv/lib/python3.10/site-packages/hcloud/hcloud.py", line 206, in _raise_exception_from_json_content
    raise APIException(
fatal: [localhost]: FAILED! => {
    "changed": false,
    "invocation": {
        "module_args": {
            "allow_deprecated_image": false,
            "api_token": "VALUE_SPECIFIED_IN_NO_LOG_PARAMETER",
            "backups": null,
            "datacenter": null,
            "delete_protection": null,
            "enable_ipv4": true,
            "enable_ipv6": true,
            "endpoint": "https://api.hetzner.cloud/v1",
            "firewalls": null,
            "force": false,
            "force_upgrade": false,
            "id": null,
            "image": "ubuntu-20.04",
            "ipv4": null,
            "ipv6": null,
            "labels": {
                "Environment": "algo"
            },
            "location": "ash",
            "name": "asdf",
            "placement_group": null,
            "rebuild_protection": null,
            "rescue_mode": null,
            "server_type": "cx11",
            "ssh_keys": [
                "(REDACTED)"
            ],
            "state": "present",
            "upgrade_disk": false,
            "user_data": "#cloud-config\noutput: {all: '| tee -a /var/log/cloud-init-output.log'}\n\npackage_update: true\npackage_upgrade: true\n\npackages:\n - sudo\n\nusers:\n  - default\n  - name: algo\n    homedir: /home/algo\n    sudo: ALL=(ALL) NOPASSWD:ALL\n    groups: adm,netdev\n    shell: /bin/bash\n    lock_passwd: true\n    ssh_authorized_keys:\n      - \"(REDACTED)\"\n\nwrite_files:\n  - path: /etc/ssh/sshd_config\n    content: |\n      Port 4160\n      AllowGroups algo\n      PermitRootLogin no\n      PasswordAuthentication no\n      ChallengeResponseAuthentication no\n      UsePAM yes\n      X11Forwarding yes\n      PrintMotd no\n      AcceptEnv LANG LC_*\n      Subsystem\tsftp\t/usr/lib/openssh/sftp-server\n\n\nruncmd:\n  - set -x\n  - ufw --force reset\n  - sudo apt-get remove -y --purge sshguard || true\n  - systemctl restart sshd.service\n",
            "volumes": null
        }
    },
    "msg": "error during placement"
}```

[quinncomendant]

Does anyone ever experience a problem, research it for an hour, find the cause, then go to GitHub to report the issue but then discover that you already created a ticket five months earlier for exactly that issue with exactly the same solution?

Because I just did that. 🤦‍♂️