Skip to content

Latest commit

 

History

History
230 lines (163 loc) · 7.87 KB

hcloud.mdx

File metadata and controls

230 lines (163 loc) · 7.87 KB
description page_title sidebar_title
The Hetzner Cloud Packer builder is able to create new images for use with the Hetzner Cloud. The builder takes a source image, runs any provisioning necessary on the image after launching it, then snapshots it into a reusable image. This reusable image can then be used as the foundation of new servers that are launched within the Hetzner Cloud.
Hetzner Cloud - Builders
Hetzner Cloud

Hetzner Cloud Builder

Type: hcloud Artifact BuilderId: hcloud.builder

The hcloud Packer builder is able to create new images for use with Hetzner Cloud. The builder takes a source image, runs any provisioning necessary on the image after launching it, then snapshots it into a reusable image. This reusable image can then be used as the foundation of new servers that are launched within the Hetzner Cloud.

The builder does not manage images. Once it creates an image, it is up to you to use it or delete it.

Connection to the server

The builder will connect to the server using the first available IP, in the following order:

  • public_ipv4: If enabled, the public IPv4 will be used,
  • public_ipv6: If enabled, the public IPv6 will be used,
  • private_ipv4: If the server is attached to private networks, the private IPv4 of the first private network will be used.

Configuration Reference

There are many configuration options available for the builder. They are segmented below into two categories: required and optional parameters. Within each category, the available configuration keys are alphabetized.

In addition to the options listed here, a communicator can be configured for this builder.

Required Builder Configuration options:

  • token (string) - The client TOKEN to use to access your account. It can also be specified via environment variable HCLOUD_TOKEN, if set.

  • image (string) - ID or name of image to launch server from. Alternatively you can use image_filter.

  • location (string) - The name of the location to launch the server in.

  • server_type (string) - ID or name of the server type this server should be created with.

Optional:

  • endpoint (string) - Non standard api endpoint URL. Set this if you are using a Hetzner Cloud API compatible service. It can also be specified via environment variable HCLOUD_ENDPOINT.

  • image_filter (object) - Filters used to populate the filter field. Example:

    image_filter {
      most_recent   = true
      with_selector = ["name==my-image"]
    }

    This selects the most recent image with the label name==my-image. NOTE: This will fail unless exactly one AMI is returned. In the above example, most_recent will cause this to succeed by selecting the newest image.

    • with_selector (list of strings) - label selectors used to select an image. NOTE: This will fail unless exactly one image is returned. Check the official hcloud docs on Label Selectors for more info.

    • most_recent (boolean) - Selects the newest created image when true. This is most useful if you base your image on another Packer build image.

    You may set this in place of image, but not both.

  • server_name (string) - The name assigned to the server. The Hetzner Cloud sets the hostname of the machine to this value.

  • server_labels (map of key/value strings) - Key/value pair labels to apply to the created server.

  • snapshot_name (string) - The name of the resulting snapshot that will appear in your account as image description. Defaults to packer-{{timestamp}} (see configuration templates for more info). The snapshot_name must be unique per architecture. If you want to reference the image as a sample in your terraform configuration please use the image id or the snapshot_labels.

  • snapshot_labels (map of key/value strings) - Key/value pair labels to apply to the created image.

  • poll_interval (string) - Configures the interval in which actions are polled by the client. Default 500ms. Increase this interval if you run into rate limiting errors.

  • user_data (string) - User data to launch with the server. Packer will not automatically wait for a user script to finish before shutting down the instance this must be handled in a provisioner.

  • user_data_file (string) - Path to a file that will be used for the user data when launching the server.

  • ssh_keys_labels (map of key/value strings) - Key/value pair labels to apply to the created ssh keys.

  • ssh_keys (array of strings) - List of SSH keys by name or id to be added to image on launch.

@include 'packer-plugin-sdk/communicator/SSHTemporaryKeyPair-not-required.mdx'

  • rescue (string) - Enable and boot in to the specified rescue system. This enables simple installation of custom operating systems. linux64 or linux32

  • upgrade_server_type (string) - ID or name of the server type this server should be upgraded to, without changing the disk size. Improves building performance. The resulting snapshot is compatible with smaller server types and disk sizes.

  • networks (array of integers) - List of Network IDs which should be attached to the server private network interface at creation time.

  • public_ipv4 (string) - ID, name or IP address of a pre-allocated Hetzner Primary IPv4 address to use for the created server.

  • public_ipv4_disabled (bool) - Disable the public ipv4 for the created server.

  • public_ipv6 (string) - ID, name or IP address of a pre-allocated Hetzner Primary IPv6 address to use for the created server.

  • public_ipv4_disabled (bool) - Disable the public ipv6 for the created server.

  • firewalls (array of strings) - List of Firewall by name or id to be attached to the created server.

Basic Example

Here is a basic example. It is completely valid as soon as you enter your own access tokens:

source "hcloud" "basic_example" {
  token = "YOUR API TOKEN"
  image = "ubuntu-22.04"
  location = "nbg1"
  server_type = "cx22"
  ssh_username = "root"
}

build {
  sources  = ["source.hcloud.basic_example"]
}

See the examples folder in the Packer project for more examples.

Tips

Keeping the images size small

To reduce the size of your images, we recommend cleaning up any temporary files that were added during the build process and discard the unused blocks from the disk.

Below are a few commands that might useful:

  • Clean and reset could-init files:

    cloud-init clean --logs --machine-id --seed --configs all
    
    rm -rf /run/cloud-init/*
    rm -rf /var/lib/cloud/*
  • Clean apt files:

    export DEBIAN_FRONTEND=noninteractive
    
    apt-get -y autopurge
    apt-get -y clean
    
    rm -rf /var/lib/apt/lists/*
  • Clean logs:

    journalctl --flush
    journalctl --rotate --vacuum-time=0
    
    find /var/log -type f -exec truncate --size 0 {} \; # truncate system logs
    find /var/log -type f -name '*.[1-9]' -delete # remove archived logs
    find /var/log -type f -name '*.gz' -delete # remove compressed archived logs
  • Reset host ssh keys:

    rm -f /etc/ssh/ssh_host_*_key /etc/ssh/ssh_host_*_key.pub

Once the cleanup is complete, you must discard the now unused blocks from the disk. This can be done with the following:

dd if=/dev/zero of=/zero bs=4M || true
sync
rm -f /zero

To speed up the process above, you may configure cloud-init to not grow the system partition to the server disk size during the boot process, leaving you with a 4GB system partition:

#cloud-config
growpart:
  mode: "off"
resize_rootfs: false

See the docker example in the Packer project for more details.