[Bug]: hcloud/setRescue: hcclient/WaitForActions: action _ failed: Unknown Error (unknown_error)

[mnencia]

What happened?

Sometimes Hetzner cloud servers fail to start.

If you run hcloud server list, you see the just created server is off.

see https://github.com/kube-hetzner/kube-hetzner/issues/49

What did you expect to happen?

The server goes up normally

Please provide a minimal working example

resource "hcloud_server" "first_control_plane" {
  name = "k3s-control-plane-0"

  image              = "ubuntu-20.04"
  rescue             = "linux64"
  server_type        = "cpx11"
  location           = "eu-central"
  ssh_keys           = [hcloud_ssh_key.k3s.id]
  firewall_ids       = [hcloud_firewall.k3s.id]
  placement_group_id = hcloud_placement_group.k3s.id

  connection {
    user           = "root"
    private_key    = local.ssh_private_key
    agent_identity = local.ssh_identity
    host           = self.ipv4_address
  }

  provisioner "file" {
    content = templatefile("${path.module}/templates/config.ign.tpl", {
      name           = self.name
      ssh_public_key = local.ssh_public_key
    })
    destination = "/root/config.ign"
  }

  provisioner "remote-exec" {
    inline = local.microOS_install_commands
  }

 ...
}

[mysticaltech]

Happens to me too many times!

[mnencia]

[mysticaltech]

I believe it fails to start sometimes when the rescue mode is requested.

[mysticaltech]

This has become a huge problem for us at https://github.com/kube-hetzner/kube-hetzner, in almost 50% of deploys it happens because we use the rescue mode to install a third-party OS on many nodes at once.

@LKaemmerling please do something about it, and any logs we can provide you, just tell us how to get them for you. Thanks!

[LKaemmerling]

Hey @mysticaltech,

would it be possible that you give me server ids where this happened?

[mysticaltech]

@LKaemmerling Here you go, this one just happened: 18212271

I will leave it in my project for you and your team to investigate more. Thanks! 🙏

[mnencia]

I just tried to provision a cluster with five nodes, and one remained off (id: 18212924)

[LKaemmerling]

Hey,

we just released v1.33.1 which contains an improvement for the situation. Can you please test it? It will be available in the next couple of minutes in the Terraform Registry.

[mysticaltech]

@LKaemmerling Thank you so much. However I have tested, and on the second try, one server, #18275324 stayed off, as before. Here are the details. Will not delete, so you can have a look if you want.

The initiating code is https://github.com/kube-hetzner/kube-hetzner

terraform --version
Terraform v1.1.6
on linux_amd64
+ provider registry.terraform.io/hashicorp/local v2.1.0
+ provider registry.terraform.io/hashicorp/null v3.1.0
+ provider registry.terraform.io/hashicorp/random v3.1.0
+ provider registry.terraform.io/hetznercloud/hcloud v1.33.1
+ provider registry.terraform.io/integrations/github v4.20.0
+ provider registry.terraform.io/tenstad/remote v0.0.23

hcloud server list after 5 minutes in:

ID         NAME                  STATUS    IPV4              IPV6                      DATACENTER
18275322   k3s-control-plane-2   running   78.46.194.108     2a01:4f8:c010:a0b1::/64   fsn1-dc14
18275323   k3s-agent-1           running   78.47.82.48       2a01:4f8:1c17:c7ac::/64   fsn1-dc14
18275324   k3s-agent-0           off       49.12.10.178      2a01:4f8:c17:8b1a::/64    fsn1-dc14
18275325   k3s-control-plane-0   running   116.202.98.33     2a01:4f8:1c17:f936::/64   fsn1-dc14
18275326   k3s-control-plane-1   running   142.132.188.100   2a01:4f8:c010:5d7f::/64   fsn1-dc14

[LKaemmerling]

Hey @mysticaltech,

your server is online ;) Just the Server status is not correct. I passed it to the specific teams. Thanks!

[mysticaltech]

In that case, fantastic! Thank you so much... :) 🙏

[apricote]

I am going to close this issue has the problem has been resolved. If this still occurs feel free to reopen the issue.