Cloud init

TL;DR
Merge 2 or more files or parts
Further readings
1. Sources

TL;DR

# Get the current status.
cloud-init status
cloud-init status --wait

# Verify that cloud-init received the expected user data.
sudo cloud-init query userdata
sudo cat /var/lib/cloud/instance/user-data.txt | gunzip

# Assert the user data we provided is a valid cloud-config.
# From version 22.2, drops the 'devel' command.
sudo cloud-init devel schema --system --annotate
sudo cloud-init devel schema --config-file '/tmp/user-data'

# Check the raw logs.
cat '/var/log/cloud-init.log'
cat '/var/log/cloud-init-output.log'

# Parse and organize the events in the log file by stage.
cloud-init analyze show

# Manually run a single cloud-config module once after the instance has booted.
sudo cloud-init single --name 'cc_ssh' --frequency 'always'

# Clean up everything so `cloud-init` can run again.
sudo cloud-init clean

# Re-run everything.
sudo cloud-init init

# Check the user scripts.
ls '/var/lib/cloud/instance/scripts'

#cloud-config

# Add the Docker repository
yum_repos:
  docker-ce:
    name: Docker CE Stable - $basearch
    enabled: true
    baseurl: https://download.docker.com/linux/rhel/$releasever/$basearch/stable
    priority: 1
    gpgcheck: true
    gpgkey: https://download.docker.com/linux/rhel/gpg

# Upgrade the instance
package_upgrade: true
package_reboot_if_required: true

# Install required packages
# docker-ce already depends on docker-ce-cli and containerd.io
packages:
  - docker-ce
  - jq
  - unzip

# Enable and start the service after installation
runcmd:
  - systemctl daemon-reload
  - systemctl enable --now docker.service

Merge 2 or more files or parts

See Merging User-Data sections for details.

#cloud-config
packages:
  - jq
  - unzip

---
merge_how:
 - name: list
   settings: [append]
 - name: dict
   settings: [no_replace, recurse_list]

packages:
  - parallel

---
packages:
  - vim

merge_type: 'list(append)+dict(recurse_array)+str()'

In Pulumi

Create a data resource containing the files in order, one per part:

new cloudinit.Config("example", {
  gzip: false,
  base64Encode: false,
  parts: [
    {
      contentType: "text/cloud-config",
      content: fs.readFileSync("cloud-init/aws.ssm.yaml", "utf8"),
      filename: "cloud-config.ssm.yml",
    },
    {
      contentType: "text/cloud-config",
      content: YAML.stringify({
        number: 3,
        plain: 'string',
        block: 'two\nlines\n'
      }),
      filename: "cloud-config.inline.yml",
      mergeType: "dict(recurse_array,no_replace)+list(append)",
    },
  ],
});

Give its rendered form as input to a virtual machine's userdata attribute, or export it:

let userData = new cloudinit.Config("userData", { … });
new aws.ec2.Instance("instance", {
  userData: userData.rendered,
  …
})

In Terraform

Create a data resource containing the files in order, one per part:

# https://registry.terraform.io/providers/hashicorp/cloudinit/latest/docs
# https://github.com/chrusty/terraform-multipart-userdata/blob/master/example/cloudinit.tf
data "cloudinit_config" "vm" {
  gzip = true
  base64_encode = true

  part {
    content      = file("${path.module}/files/cloud-init/first.yaml")
    content_type = "text/cloud-config"
    filename     = "first.yaml"
  }
  …
  part {
    content      = templatefile(
      "${path.module}/templates/cloud-init/n-th.yaml.tftpl",
      {
        key = value
      }
    )
    content_type = "text/cloud-config"
    merge_type   = "dict(recurse_array,no_replace)+list(append)"
    filename     = "n-th.yaml"
  }
}

Give its rendered form as input to a virtual machine's userdata attribute or an output resource:

resource "azurerm_linux_virtual_machine" "vm" {
  user_data = data.cloudinit_config.vm.rendered
  …
}

output "cloudinit_config" {
  value = data.cloudinit_config.vm.rendered
}

5.9 KiB Raw Blame History

Cloud init

Table of contents

TL;DR

Merge 2 or more files or parts

Further readings

Sources

5.9 KiB

Raw Blame History