Merge pull request #103 from kube-hetzner/cloud-init2

Switch to cloud-init for host initialization

.gitignore

@@ -4,5 +4,11 @@ crash.log
 kubeconfig.yaml
 kubeconfig.yaml-e
 terraform.tfvars
+<<<<<<< HEAD
+plans-custom.yaml
+traefik-custom.yaml
+kured-custom.yaml
+=======
 plans.yaml
 traefik_config.yaml
+>>>>>>> 32d43b50f40b30cba207fb5946dee422825d93ef

agents.tf

@@ -44,11 +44,11 @@ resource "null_resource" "agents" {
   provisioner "file" {
     content = yamlencode({
       node-name     = module.agents[each.key].name
-      server        = "https://${local.first_control_plane_network_ipv4}:6443"
+      server        = "https://${module.control_planes[0].private_ipv4_address}:6443"
       token         = random_password.k3s_token.result
       kubelet-arg   = "cloud-provider=external"
       flannel-iface = "eth1"
-      node-ip       = module.agents[each.key].ipv4_address
+      node-ip       = module.agents[each.key].private_ipv4_address
       node-label    = var.automatically_upgrade_k3s ? ["k3s_upgrade=true"] : []
     })
     destination = "/tmp/config.yaml"

control_planes.tf

@@ -53,7 +53,6 @@ resource "null_resource" "control_planes" {
       kubelet-arg              = "cloud-provider=external"
       node-ip                  = module.control_planes[count.index].private_ipv4_address
       advertise-address        = module.control_planes[count.index].private_ipv4_address
-      tls-san                  = module.control_planes[count.index].private_ipv4_address
       node-taint               = var.allow_scheduling_on_control_plane ? [] : ["node-role.kubernetes.io/master:NoSchedule"]
       node-label               = var.automatically_upgrade_k3s ? ["k3s_upgrade=true"] : []
     })

init.tf

@@ -18,7 +18,6 @@ resource "null_resource" "first_control_plane" {
       kubelet-arg              = "cloud-provider=external"
       node-ip                  = module.control_planes[0].private_ipv4_address
       advertise-address        = module.control_planes[0].private_ipv4_address
-      tls-san                  = module.control_planes[0].private_ipv4_address
       node-taint               = var.allow_scheduling_on_control_plane ? [] : ["node-role.kubernetes.io/master:NoSchedule"]
       node-label               = var.automatically_upgrade_k3s ? ["k3s_upgrade=true"] : []
     })
@@ -30,7 +29,7 @@ resource "null_resource" "first_control_plane" {
     inline = local.install_k3s_server
   }
 
-  # Upon reboot verify that the k3s server is starts, and wait for k3s to be ready to receive commands
+  # Upon reboot start k3s and wait for it to be ready to receive commands
   provisioner "remote-exec" {
     inline = [
       "systemctl start k3s",

locals.tf

@@ -1,6 +1,4 @@
 locals {
-  first_control_plane_network_ipv4 = module.control_planes[0].private_ipv4_address
-
   ssh_public_key = trimspace(file(var.public_key))
   # ssh_private_key is either the contents of var.private_key or null to use a ssh agent.
   ssh_private_key = var.private_key == null ? null : trimspace(file(var.private_key))
@@ -145,9 +143,10 @@ locals {
     "[ -e /etc/rancher/k3s/k3s.yaml ] && exit 0",
   ]
 
-  install_k3s_server = concat(local.common_commands_install_k3s, ["curl -sfL https://get.k3s.io | INSTALL_K3S_SKIP_SELINUX_RPM=true INSTALL_K3S_SKIP_START=true INSTALL_K3S_CHANNEL=${var.initial_k3s_channel} INSTALL_K3S_EXEC=server sh -"])
+  apply_k3s_selinux = ["/sbin/semodule -v -i /usr/share/selinux/packages/k3s.pp"]
 
-  install_k3s_agent = concat(local.common_commands_install_k3s, ["curl -sfL https://get.k3s.io | INSTALL_K3S_SKIP_SELINUX_RPM=true INSTALL_K3S_SKIP_START=true INSTALL_K3S_CHANNEL=${var.initial_k3s_channel} INSTALL_K3S_EXEC=agent sh -"])
+  install_k3s_server = concat(local.common_commands_install_k3s, ["curl -sfL https://get.k3s.io | INSTALL_K3S_SKIP_START=true INSTALL_K3S_SKIP_SELINUX_RPM=true INSTALL_K3S_CHANNEL=${var.initial_k3s_channel} INSTALL_K3S_EXEC=server sh -"], local.apply_k3s_selinux)
+  install_k3s_agent  = concat(local.common_commands_install_k3s, ["curl -sfL https://get.k3s.io | INSTALL_K3S_SKIP_START=true INSTALL_K3S_SKIP_SELINUX_RPM=true INSTALL_K3S_CHANNEL=${var.initial_k3s_channel} INSTALL_K3S_EXEC=agent sh -"], local.apply_k3s_selinux)
 
   agent_nodepools = merge([
     for nodepool_name, nodepool_obj in var.agent_nodepools : {

modules/host/locals.tf

@@ -10,65 +10,4 @@ locals {
   ssh_identity_file = var.private_key == null ? var.public_key : var.private_key
   # shared flags for ssh to ignore host keys, to use our ssh identity file for all connections during provisioning.
   ssh_args = "-o UserKnownHostsFile=/dev/null -o StrictHostKeyChecking=no -i ${local.ssh_identity_file}"
-
-  microOS_install_commands = [
-    "set -ex",
-    "apt-get update",
-    "apt-get install -y aria2",
-    "aria2c --follow-metalink=mem https://download.opensuse.org/tumbleweed/appliances/openSUSE-MicroOS.x86_64-kvm-and-xen.qcow2.meta4",
-    "qemu-img convert -p -f qcow2 -O host_device $(ls -a | grep -ie '^opensuse.*microos.*qcow2$') /dev/sda",
-    "sgdisk -e /dev/sda",
-    "parted -s /dev/sda resizepart 4 99%",
-    "parted -s /dev/sda mkpart primary ext2 99% 100%",
-    "partprobe /dev/sda && udevadm settle && fdisk -l /dev/sda",
-    "mount /dev/sda4 /mnt/ && btrfs filesystem resize max /mnt && umount /mnt",
-    "mke2fs -L ignition /dev/sda5",
-    "mount /dev/sda5 /mnt",
-    "mkdir /mnt/ignition",
-    "cp /root/config.ign /mnt/ignition/config.ign",
-    "mkdir /mnt/combustion",
-    "cp /root/script /mnt/combustion/script",
-    "umount /mnt"
-  ]
-
-  ignition_config = jsonencode({
-    ignition = {
-      version = "3.0.0"
-    }
-    passwd = {
-      users = [{
-        name              = "root"
-        sshAuthorizedKeys = concat([local.ssh_public_key], var.additional_public_keys)
-      }]
-    }
-    storage = {
-      files = [
-        {
-          path      = "/etc/sysconfig/network/ifcfg-eth1"
-          mode      = 420
-          overwrite = true
-          contents  = { "source" = "data:,BOOTPROTO%3D%27dhcp%27%0ASTARTMODE%3D%27auto%27" }
-        },
-        {
-          path      = "/etc/ssh/sshd_config.d/kube-hetzner.conf"
-          mode      = 420
-          overwrite = true
-          contents  = { "source" = "data:,PasswordAuthentication%20no%0AX11Forwarding%20no%0AMaxAuthTries%202%0AAllowTcpForwarding%20no%0AAllowAgentForwarding%20no%0AAuthorizedKeysFile%20.ssh%2Fauthorized_keys" }
-        }
-      ]
-    }
-  })
-
-  combustion_script = <<EOF
-#!/bin/bash
-sed -i 's#NETCONFIG_NIS_SETDOMAINNAME="yes"#NETCONFIG_NIS_SETDOMAINNAME="no"#g' /etc/sysconfig/network/config
-sed -i 's#WAIT_FOR_INTERFACES="30"#WAIT_FOR_INTERFACES="60"#g' /etc/sysconfig/network/config
-sed -i 's#CHECK_DUPLICATE_IP="yes"#CHECK_DUPLICATE_IP="no"#g' /etc/sysconfig/network/config
-# combustion: network
-rpm --import https://rpm.rancher.io/public.key
-zypper refresh
-zypper --gpg-auto-import-keys install -y https://rpm.rancher.io/k3s/stable/common/microos/noarch/k3s-selinux-0.4-1.sle.noarch.rpm
-udevadm settle || true
-EOF
-
 }

modules/host/main.tf

@@ -8,10 +8,20 @@ resource "hcloud_server" "server" {
   ssh_keys           = var.ssh_keys
   firewall_ids       = var.firewall_ids
   placement_group_id = var.placement_group_id
-
+  user_data          = data.template_cloudinit_config.config.rendered
 
   labels = var.labels
 
+  # Prevent destroying the whole cluster if the user changes
+  # any of the attributes that force to recreate the servers.
+  lifecycle {
+    ignore_changes = [
+      location,
+      ssh_keys,
+      user_data,
+    ]
+  }
+
   connection {
     user           = "root"
     private_key    = local.ssh_private_key
@@ -19,30 +29,21 @@ resource "hcloud_server" "server" {
     host           = self.ipv4_address
   }
 
-  provisioner "file" {
-    content     = local.ignition_config
-    destination = "/root/config.ign"
-  }
-
-  # Combustion script file to install k3s-selinux
-  provisioner "file" {
-    content     = local.combustion_script
-    destination = "/root/script"
-  }
-
   # Install MicroOS
   provisioner "remote-exec" {
-    inline = local.microOS_install_commands
+    inline = [
+      "set -ex",
+      "apt-get update",
+      "apt-get install -y aria2",
+      "aria2c --follow-metalink=mem https://download.opensuse.org/tumbleweed/appliances/openSUSE-MicroOS.x86_64-OpenStack-Cloud.qcow2.meta4",
+      "qemu-img convert -p -f qcow2 -O host_device $(ls -a | grep -ie '^opensuse.*microos.*qcow2$') /dev/sda",
+    ]
   }
 
-  # Issue a reboot command
+  # Issue a reboot command and wait for MicroOS to reboot and be ready
-  provisioner "local-exec" {
-    command = "ssh ${local.ssh_args} root@${self.ipv4_address} '(sleep 2; reboot)&'; sleep 3"
-  }
-
-  # Wait for MicroOS to reboot and be ready
   provisioner "local-exec" {
     command = <<-EOT
+      ssh ${local.ssh_args} root@${self.ipv4_address} '(sleep 2; reboot)&'; sleep 3
       until ssh ${local.ssh_args} -o ConnectTimeout=2 root@${self.ipv4_address} true 2> /dev/null
       do
         echo "Waiting for MicroOS to reboot and become available..."
@@ -51,16 +52,25 @@ resource "hcloud_server" "server" {
     EOT
   }
 
+  # Install k3s-selinux (compatible version)
   provisioner "remote-exec" {
     inline = [
-      # Disable automatic reboot (after transactional updates), and configure the reboot method as kured
       "set -ex",
-      "rebootmgrctl set-strategy off",
+      "transactional-update pkg install -y k3s-selinux"
-      "echo 'REBOOT_METHOD=kured' > /etc/transactional-update.conf",
-      # set the hostname
-      "hostnamectl set-hostname ${self.name}"
     ]
   }
+
+  # Issue a reboot command and wait for MicroOS to reboot and be ready
+  provisioner "local-exec" {
+    command = <<-EOT
+      ssh ${local.ssh_args} root@${self.ipv4_address} '(sleep 2; reboot)&'; sleep 3
+      until ssh ${local.ssh_args} -o ConnectTimeout=2 root@${self.ipv4_address} true 2> /dev/null
+      do
+        echo "Waiting for MicroOS to reboot and become available..."
+        sleep 3
+      done
+    EOT
+  }
 }
 
 resource "hcloud_server_network" "server" {
@@ -68,3 +78,21 @@ resource "hcloud_server_network" "server" {
   server_id = hcloud_server.server.id
   subnet_id = var.ipv4_subnet_id
 }
+
+data "template_cloudinit_config" "config" {
+  gzip          = true
+  base64_encode = true
+
+  # Main cloud-config configuration file.
+  part {
+    filename     = "init.cfg"
+    content_type = "text/cloud-config"
+    content = templatefile(
+      "${path.module}/templates/userdata.yaml.tpl",
+      {
+        hostname          = var.name
+        sshAuthorizedKeys = concat([local.ssh_public_key], var.additional_public_keys)
+      }
+    )
+  }
+}

modules/host/templates/userdata.yaml.tpl

@@ -0,0 +1,58 @@
+#cloud-config
+
+write_files:
+
+# Configure the private network interface
+- content: |
+    BOOTPROTO='dhcp'
+    STARTMODE='auto'
+  path: /etc/sysconfig/network/ifcfg-eth1
+
+# Disable ssh password authentication
+- content: |
+    PasswordAuthentication no
+    X11Forwarding no
+    MaxAuthTries 2
+    AllowTcpForwarding no
+    AllowAgentForwarding no
+    AuthorizedKeysFile .ssh/authorized_keys
+  path: /etc/ssh/sshd_config.d/kube-hetzner.conf
+
+# Set reboot method as "kured"
+- content: |
+    REBOOT_METHOD=kured
+  path: /etc/transactional-update.conf
+
+# Add ssh authorized keys
+ssh_authorized_keys:
+%{ for key in sshAuthorizedKeys ~}
+  - ${key}
+%{ endfor ~}
+
+# Resize /var, not /, as that's the last partition in MicroOS image.
+growpart:
+    devices: ["/var"]
+
+# Make sure the hostname is set correctly
+hostname: ${hostname}
+preserve_hostname: true
+
+runcmd:
+
+# As above, make sure the hostname is not reset
+- [sed, '-i', 's/NETCONFIG_NIS_SETDOMAINNAME="yes"/NETCONFIG_NIS_SETDOMAINNAME="no"/g', /etc/sysconfig/network/config]
+- [sed, '-i', 's/DHCLIENT_SET_HOSTNAME="yes"/DHCLIENT_SET_HOSTNAME="no"/g', /etc/sysconfig/network/dhcp]
+
+# We set Cloudflare DNS servers, followed by Google as a backup
+- [sed, '-i', 's/NETCONFIG_DNS_STATIC_SERVERS=""/NETCONFIG_DNS_STATIC_SERVERS="1.1.1.1 1.0.0.1 8.8.8.8"/g', /etc/sysconfig/network/config]
+
+# Bounds the amount of logs that can survive on the system
+- [sed, '-i', 's/#SystemMaxUse=/SystemMaxUse=3G/g', /etc/systemd/journald.conf]
+- [sed, '-i', 's/#MaxRetentionSec=/MaxRetentionSec=1week/g', /etc/systemd/journald.conf]
+
+# Reduces the default number of snapshots from 2-10 number limit, to 4 and from 4-10 number limit important, to 2
+- [sed, '-i', 's/NUMBER_LIMIT="2-10"/NUMBER_LIMIT="4"/g', /etc/snapper/configs/root]
+- [sed, '-i', 's/NUMBER_LIMIT_IMPORTANT="4-10"/NUMBER_LIMIT_IMPORTANT="3"/g', /etc/snapper/configs/root]
+
+# Disables unneeded services
+- [systemctl, disable, '--now', 'rebootmgr.service']

modules/host/versions.tf

@@ -12,5 +12,9 @@ terraform {
       source  = "tenstad/remote"
       version = "~> 0.0.23"
     }
+    template = {
+      source  = "hashicorp/template"
+      version = "~> 2.2.0"
+    }
   }
 }

terraform.tfvars.example

@@ -1,7 +1,15 @@
-# You need to replace these
+# Only the first values starting with a * are obligatory, the rest can remain with their default values, or you
+# could adapt them to your needs.
+#
+# Note that some values, notably "location" and "public_key" have no effect after the initial cluster has been setup.
+# This is in order to keep terraform from re-provisioning all nodes at once which would loose data. If you want to update,
+# those, you should instead change the value here and then manually re-provision each node one-by-one. Grep for "lifecycle".
+
+# * Your Hetzner project API token 
 hcloud_token = "xxxxxxxxxxxxxxxxxxYYYYYYYYYYYYYYYYYYYzzzzzzzzzzzzzzzzzzzzz"
+# * Your public key
 public_key   = "/home/username/.ssh/id_ed25519.pub"
-# Must be "private_key = null" when you want to use ssh-agent, for a Yubikey like device auth or an SSH key-pair with passphrase
+# * Your private key, must be "private_key = null" when you want to use ssh-agent, for a Yubikey like device auth or an SSH key-pair with passphrase
 private_key  = "/home/username/.ssh/id_ed25519"
 
 # These can be customized, or left with the default values
@@ -10,9 +18,6 @@ private_key  = "/home/username/.ssh/id_ed25519"
 location         = "fsn1"       # change to `ash` for us-east Ashburn, Virginia location
 network_region   = "eu-central" # change to `us-east` if location is ash
 
-# It's best to leave the network range as is, unless you know what you are doing. The default is "10.0.0.0/8".
-# network_ipv4_range = "10.0.0.0/8"
-
 # You can have up to as many subnets as you want (preferably if the form of 10.X.0.0/16),
 # their primary use is to logically separate the nodes.
 # The control_plane network is mandatory.
@@ -48,6 +53,11 @@ agent_nodepools = {
 # That will depend on how much load you want it to handle, see https://www.hetzner.com/cloud/load-balancer
 load_balancer_type        = "lb11"
 
+### The following values are fully optional
+
+# It's best to leave the network range as is, unless you know what you are doing. The default is "10.0.0.0/8".
+# network_ipv4_range = "10.0.0.0/8"
+
 # If you want to use a specific Hetzner CCM and CSI version, set them below, otherwise leave as is for the latest versions
 # hetzner_ccm_version = ""
 # hetzner_csi_version = ""