cloud-init Documentation

Cloud-init is the industry standard multi-distribution method for cross-platform cloud instance initialization. It is supported across all major public cloud providers, provisioning systems for private cloud infrastructure, and bare-metal installations.

Cloud instances are initialized from a disk image and instance data:

• Cloud metadata

• User data (optional)

• Vendor data (optional)

Cloud-init will identify the cloud it is running on during boot, read any provided metadata from the cloud and initialize the system accordingly. This may involve setting up the network and storage devices to configuring SSH access key and many other aspects of a system. Later on the cloud-init will also parse and process any optional user or vendor data that was passed to the instance.

Getting help

Having trouble? We would like to help!

• Try the FAQ – its got answers to some common questions

• Ask a question in the #cloud-init IRC channel on Libera

• Join and ask questions on the cloud-init mailing list

• Find a bug? Report bugs on Launchpad

Availability

Below outlines the current availability of cloud-init across distributions and clouds, both public and private.

Note

If a distribution or cloud does not show up in the list below contact them and ask for images to be generated using cloud-init!

Distributions

Cloud-init has support across all major Linux distributions, FreeBSD, NetBSD, OpenBSD and DragonFlyBSD:

• Alpine Linux

• ArchLinux

• Debian

• DragonFlyBSD

• Fedora

• FreeBSD

• Gentoo Linux

• NetBSD

• OpenBSD

• Photon OS

• RHEL/CentOS/AlmaLinux/Rocky Linux/EuroLinux

• SLES/openSUSE

• Ubuntu

Clouds

Cloud-init provides support across a wide ranging list of execution environments in the public cloud:

• Amazon Web Services

• Microsoft Azure

• Google Cloud Platform

• Oracle Cloud Infrastructure

• Softlayer

• Rackspace Public Cloud

• IBM Cloud

• DigitalOcean

• Bigstep

• Hetzner

• Joyent

• CloudSigma

• Alibaba Cloud

• OVH

• OpenNebula

• Exoscale

• Scaleway

• CloudStack

• AltCloud

• SmartOS

• UpCloud

• Vultr

• Zadara Edge Cloud Platform

Additionally, cloud-init is supported on these private clouds:

• Bare metal installs

• OpenStack

• LXD

• KVM

• Metal-as-a-Service (MAAS)

• VMware

Boot Stages

In order to be able to provide the functionality that it does, cloud-init must be integrated into the boot in fairly controlled way. There are five stages to boot:

1. Generator

2. Local

3. Network

4. Config

5. Final

Generator

When booting under systemd, a generator will run that determines if cloud-init.target should be included in the boot goals. By default, this generator will enable cloud-init. It will not enable cloud-init if either:

• The file /etc/cloud/cloud-init.disabled exists

• The kernel command line as found in /proc/cmdline contains cloud-init=disabled. When running in a container, the kernel command line is not honored, but cloud-init will read an environment variable named KERNEL_CMDLINE in its place.

Again, these mechanisms for disabling cloud-init at runtime currently only exist in systemd.

Local

systemd service		cloud-init-local.service
runs		as soon as possible with / mounted read-write
blocks		as much of boot as possible, must block network
modules		none

The purpose of the local stage is to:

• locate “local” data sources.

• apply networking configuration to the system (including “Fallback”)

In most cases, this stage does not do much more than that. It finds the datasource and determines the network configuration to be used. That network configuration can come from:

• datasource: cloud provided network configuration via metadata

• fallback: cloud-init’s fallback networking consists of rendering the equivalent to “dhcp on eth0”, which was historically the most popular mechanism for network configuration of a guest

• none: network configuration can be disabled by writing the file /etc/cloud/cloud.cfg with the content: network: {config: disabled}

If this is an instance’s first boot, then the selected network configuration is rendered. This includes clearing of all previous (stale) configuration including persistent device naming with old mac addresses.

This stage must block network bring-up or any stale configuration that might have already been applied. Otherwise, that could have negative effects such as DHCP hooks or broadcast of an old hostname. It would also put the system in an odd state to recover from, as it may then have to restart network devices.

Cloud-init then exits and expects for the continued boot of the operating system to bring network configuration up as configured.

Note: In the past, local data sources have been only those that were available without network (such as ‘ConfigDrive’). However, as seen in the recent additions to the DigitalOcean datasource, even data sources that require a network can operate at this stage.

Network

systemd service		cloud-init.service
runs		after local stage and configured networking is up
blocks		as much of remaining boot as possible
modules		cloud_init_modules in /etc/cloud/cloud.cfg

This stage requires all configured networking to be online, as it will fully process any user-data that is found. Here processing means:

• retrieve any #include or #include-once (recursively) including http

• decompress any compressed content

• run any part-handler found.

This stage runs the disk_setup and mounts modules which may partition and format disks and configure mount points (such as in /etc/fstab). Those modules cannot run earlier as they may receive configuration input from sources only available via network. For example, a user may have provided user-data in a network resource that describes how local mounts should be done.

On some clouds, such as Azure, this stage will create filesystems to be mounted, including ones that have stale (previous instance) references in /etc/fstab. As such, entries /etc/fstab other than those necessary for cloud-init to run should not be done until after this stage.

A part-handler will run at this stage, as will boot-hooks including cloud-config bootcmd. The user of this functionality has to be aware that the system is in the process of booting when their code runs.

Config

systemd service		cloud-config.service
runs		after network
blocks		nothing
modules		cloud_config_modules in /etc/cloud/cloud.cfg

This stage runs config modules only. Modules that do not really have an effect on other stages of boot are run here, including runcmd.

Final

systemd service		cloud-final.service
runs		as final part of boot (traditional “rc.local”)
blocks		nothing
modules		cloud_final_modules in /etc/cloud/cloud.cfg

This stage runs as late in boot as possible. Any scripts that a user is accustomed to running after logging into a system should run correctly here. Things that run here include:

• package installations

• configuration management plugins (puppet, chef, salt-minion)

• user-defined scripts (i.e. shell scripts passed as user-data)

For scripts external to cloud-init looking to wait until cloud-init is finished, the cloud-init status subcommand can help block external scripts until cloud-init is done without having to write your own systemd units dependency chains. See status for more info.

First Boot Determination

cloud-init has to determine whether or not the current boot is the first boot of a new instance or not, so that it applies the appropriate configuration. On an instance’s first boot, it should run all “per-instance” configuration, whereas on a subsequent boot it should run only “per-boot” configuration. This section describes how cloud-init performs this determination, as well as why it is necessary.

When it runs, cloud-init stores a cache of its internal state for use across stages and boots.

If this cache is present, then cloud-init has run on this system before. 1 There are two cases where this could occur. Most commonly, the instance has been rebooted, and this is a second/subsequent boot. Alternatively, the filesystem has been attached to a new instance, and this is an instance’s first boot. The most obvious case where this happens is when an instance is launched from an image captured from a launched instance.

By default, cloud-init attempts to determine which case it is running in by checking the instance ID in the cache against the instance ID it determines at runtime. If they do not match, then this is an instance’s first boot; otherwise, it’s a subsequent boot. Internally, cloud-init refers to this behavior as check.

This behavior is required for images captured from launched instances to behave correctly, and so is the default which generic cloud images ship with. However, there are cases where it can cause problems. 2 For these cases, cloud-init has support for modifying its behavior to trust the instance ID that is present in the system unconditionally. This means that cloud-init will never detect a new instance when the cache is present, and it follows that the only way to cause cloud-init to detect a new instance (and therefore its first boot) is to manually remove cloud-init’s cache. Internally, this behavior is referred to as trust.

To configure which of these behaviors to use, cloud-init exposes the manual_cache_clean configuration option. When false (the default), cloud-init will check and clean the cache if the instance IDs do not match (this is the default, as discussed above). When true, cloud-init will trust the existing cache (and therefore not clean it).

Manual Cache Cleaning

cloud-init ships a command for manually cleaning the cache: cloud-init clean. See clean’s documentation for further details.

Reverting manual_cache_clean Setting

Currently there is no support for switching an instance that is launched with manual_cache_clean: true from trust behavior to check behavior, other than manually cleaning the cache.

Warning

If you want to capture an instance that is currently in trust mode as an image for launching other instances, you must manually clean the cache. If you do not do so, then instances launched from the captured image will all detect their first boot as a subsequent boot of the captured instance, and will not apply any per-instance configuration.

This is a functional issue, but also a potential security one: cloud-init is responsible for rotating SSH host keys on first boot, and this will not happen on these instances.

1

It follows that if this cache is not present, cloud-init has not run on this system before, so this is unambiguously this instance’s first boot.

2

A couple of ways in which this strict reliance on the presence of a datasource has been observed to cause problems:

• If a cloud’s metadata service is flaky and cloud-init cannot obtain the instance ID locally on that platform, cloud-init’s instance ID determination will sometimes fail to determine the current instance ID, which makes it impossible to determine if this is an instance’s first or subsequent boot (#1885527).

• If cloud-init is used to provision a physical appliance or device and an attacker can present a datasource to the device with a different instance ID, then cloud-init’s default behavior will detect this as an instance’s first boot and reset the device using the attacker’s configuration (this has been observed with the NoCloud datasource in #1879530).

CLI Interface

For the latest list of subcommands and arguments use cloud-init’s --help option. This can be used against cloud-init itself or any of its subcommands.

$ cloud-init --help
  usage: /usr/bin/cloud-init [-h] [--version] [--file FILES] [--debug] [--force]
                          {init,modules,single,query,dhclient-hook,features,analyze,devel,collect-logs,clean,status}
                          ...

  optional arguments:
  -h, --help            show this help message and exit
  --version, -v         show program's version number and exit
  --file FILES, -f FILES
                        additional yaml configuration files to use
  --debug, -d           show additional pre-action logging (default: False)
  --force               force running even if no datasource is found (use at
                        your own risk)

  Subcommands:
  {init,modules,single,query,dhclient-hook,features,analyze,devel,collect-logs,clean,status}
      init                initializes cloud-init and performs initial modules
      modules             activates modules using a given configuration key
      single              run a single module
      query               Query standardized instance metadata from the command
                          line.
      dhclient-hook       Run the dhclient hook to record network info.
      features            list defined features
      analyze             Devel tool: Analyze cloud-init logs and data
      devel               Run development tools
      collect-logs        Collect and tar all cloud-init debug info
      clean               Remove logs and artifacts so cloud-init can re-run.
      status              Report cloud-init status or wait on completion.

The rest of this document will give an overview of each of the subcommands.

analyze

Get detailed reports of where cloud-init spends its time during the boot process. For more complete reference see analyze.

Possible subcommands include:

• blame: report ordered by most costly operations

• dump: machine-readable JSON dump of all cloud-init tracked events

• show: show time-ordered report of the cost of operations during each boot stage

• boot: show timestamps from kernel initialization, kernel finish initialization, and cloud-init start

clean

Remove cloud-init artifacts from /var/lib/cloud to simulate a clean instance. On reboot, cloud-init will re-run all stages as it did on first boot.

• --logs: optionally remove all cloud-init log files in /var/log/

• --reboot: reboot the system after removing artifacts

collect-logs

Collect and tar cloud-init generated logs, data files, and system information for triage. This subcommand is integrated with apport.

Logs collected include:

• /var/log/cloud-init.log

• /var/log/cloud-init-output.log

• /run/cloud-init

• /var/lib/cloud/instance/user-data.txt

• cloud-init package version

• dmesg output

• journalctl output

Note

Ubuntu users can file bugs with ubuntu-bug cloud-init to automatically attach these logs to a bug report

devel

Collection of development tools under active development. These tools will likely be promoted to top-level subcommands when stable.

Do NOT rely on the output of these commands as they can and will change.

Current subcommands:

• net-convert: manually use cloud-init’s network format conversion, useful for testing configuration or testing changes to the network conversion logic itself.

• render: use cloud-init’s jinja template render to process #cloud-config or custom-scripts, injecting any variables from /run/cloud-init/instance-data.json. It accepts a user-data file containing the jinja template header ## template: jinja and renders that content with any instance-data.json variables present.

• schema: a #cloud-config format and schema validator. It accepts a cloud-config YAML file and annotates potential schema errors locally without the need for deployment. Schema validation is work in progress and supports a subset of cloud-config modules.

• hotplug-hook: respond to newly added system devices by retrieving updated system metadata and bringing up/down the corresponding device. This command is intended to be called via a systemd service and is not considered user-accessible except for debugging purposes.

features

Print out each feature supported. If cloud-init does not have the features subcommand, it also does not support any features described in this document.

$ cloud-init features
NETWORK_CONFIG_V1
NETWORK_CONFIG_V2

init

Generally run by OS init systems to execute cloud-init’s stages init and init-local. See Boot Stages for more info. Can be run on the commandline, but is generally gated to run only once due to semaphores in /var/lib/cloud/instance/sem/ and /var/lib/cloud/sem.

• --local: run init-local stage instead of init

modules

Generally run by OS init systems to execute modules:config and modules:final boot stages. This executes cloud config Modules configured to run in the init, config and final stages. The modules are declared to run in various boot stages in the file /etc/cloud/cloud.cfg under keys:

• cloud_init_modules

• cloud_config_modules

• cloud_final_modules

Can be run on the command line, but each module is gated to run only once due to semaphores in /var/lib/cloud/.

• --mode [init|config|final]: run modules:init, modules:config or modules:final cloud-init stages. See Boot Stages for more info.

query

Query standardized cloud instance metadata crawled by cloud-init and stored in /run/cloud-init/instance-data.json. This is a convenience command-line interface to reference any cached configuration metadata that cloud-init crawls when booting the instance. See Instance Metadata for more info.

• --all: dump all available instance data as json which can be queried

• --instance-data: optional path to a different instance-data.json file to source for queries

• --list-keys: list available query keys from cached instance data

• --format: a string that will use jinja-template syntax to render a string replacing

• <varname>: a dot-delimited variable path into the instance-data.json object

Below demonstrates how to list all top-level query keys that are standardized aliases:

$ cloud-init query --list-keys
_beta_keys
availability_zone
base64_encoded_keys
cloud_name
ds
instance_id
local_hostname
platform
public_ssh_keys
region
sensitive_keys
subplatform
userdata
v1
vendordata

Below demonstrates how to query standardized metadata from clouds:

% cloud-init query v1.cloud_name
aws  # or openstack, azure, gce etc.

# Any standardized instance-data under a <v#> key is aliased as a top-level key for convenience.
% cloud-init query cloud_name
aws  # or openstack, azure, gce etc.

# Query datasource-specific metadata on EC2
% cloud-init query ds.meta_data.public_ipv4

Note

The standardized instance data keys under v# are guaranteed not to change behavior or format. If using top-level convenience aliases for any standardized instance data keys, the most value (highest v#) of that key name is what is reported as the top-level value. So these aliases act as a ‘latest’.

This data can then be formatted to generate custom strings or data:

# Generate a custom hostname fqdn based on instance-id, cloud and region
% cloud-init query --format 'custom-{{instance_id}}.{{region}}.{{v1.cloud_name}}.com'
custom-i-0e91f69987f37ec74.us-east-2.aws.com

single

Attempt to run a single named cloud config module.

• --name: the cloud-config module name to run

• --frequency: optionally override the declared module frequency with one of (always|once-per-instance|once)

The following example re-runs the cc_set_hostname module ignoring the module default frequency of once-per-instance:

$ cloud-init single --name set_hostname --frequency always

Note

Mileage may vary trying to re-run each cloud-config module, as some are not idempotent.

status

Report whether cloud-init is running, done, disabled or errored. Exits non-zero if an error is detected in cloud-init.

• --long: detailed status information

• --wait: block until cloud-init completes

Below are examples of output when cloud-init is running, showing status and the currently running modules, as well as when it is done.

$ cloud-init status
status: running

$ cloud-init status --long
status: running
time: Fri, 26 Jan 2018 21:39:43 +0000
detail:
Running in stage: init-local

$ cloud-init status
status: done

$ cloud-init status --long
status: done
time: Wed, 17 Jan 2018 20:41:59 +0000
detail:
DataSourceNoCloud [seed=/var/lib/cloud/seed/nocloud-net][dsmode=net]

FAQ

How do I get help?

Having trouble? We would like to help!

• First go through this page with answers to common questions

• Use the search bar at the upper left to search these docs

• Ask a question in the #cloud-init IRC channel on Libera

• Join and ask questions on the cloud-init mailing list

• Find a bug? Check out the Reporting Bugs topic for how to report one

Where are the logs?

Cloud-init uses two files to log to:

• /var/log/cloud-init-output.log: captures the output from each stage of cloud-init when it runs

• /var/log/cloud-init.log: very detailed log with debugging output, detailing each action taken

• /run/cloud-init: contains logs about how cloud-init decided to enable or disable itself, as well as what platforms/datasources were detected. These logs are most useful when trying to determine what cloud-init ran or did not run.

Be aware that each time a system boots, new logs are appended to the files in /var/log. Therefore, the files may have more than one boot worth of information present.

When reviewing these logs look for any errors or Python tracebacks to check for any errors.

Where are the configuration files?

Cloud-init config is provided in two places:

• /etc/cloud/cloud.cfg

• /etc/cloud/cloud.cfg.d/*.cfg

These files can define the modules that run during instance initialization, the datasources to evaluate on boot, and other settings.

Where are the data files?

Inside the /var/lib/cloud/ directory there are two important subdirectories:

instance

The /var/lib/cloud/instance directory is a symbolic link that points to the most recently used instance-id directory. This folder contains the information cloud-init received from datasources, including vendor and user data. This can be helpful to review to ensure the correct data was passed.

It also contains the datasource file that containers the full information about what datasource was identified and used to setup the system.

Finally, the boot-finished file is the last thing that cloud-init does.

data

The /var/lib/cloud/data directory contain information related to the previous boot:

• instance-id: id of the instance as discovered by cloud-init. Changing this file has no effect.

• result.json: json file will show both the datasource used to setup the instance, and if any errors occurred

• status.json: json file shows the datasource used and a break down of all four modules if any errors occurred and the start and stop times.

What datasource am I using?

To correctly setup an instance, cloud-init must correctly identify the cloud that it is on. Therefore knowing what datasource is used on an instance launch can help aid in debugging.

To find what datasource is getting used run the cloud-id command:

$ cloud-id
nocloud

If the cloud-id is not what is expected, then running the ds-identify script in debug mode and providing that in a bug can help aid in resolving any issues:

$ sudo DEBUG_LEVEL=2 DI_LOG=stderr /usr/lib/cloud-init/ds-identify --force

The force parameter allows the command to be run again since the instance has already launched. The other options increase the verbosity of logging and put the logs to STDERR.

How can I re-run datasource detection and cloud-init?

If a user is developing a new datasource or working on debugging an issue it may be useful to re-run datasource detection and the initial setup of cloud-init.

To do this, force ds-identify to re-run, clean up any logs, and re-run cloud-init:

$ sudo DI_LOG=stderr /usr/lib/cloud-init/ds-identify --force
$ sudo cloud-init clean --logs
$ sudo cloud-init init --local
$ sudo cloud-init init

Warning

These commands will re-run cloud-init as if this were first boot of a system: this will, at the very least, cycle SSH host keys and may do substantially more. Do not run these commands on production systems.

How can I debug my user data?

Two of the most common issues with user data, that also happens to be cloud-config is:

1. Incorrectly formatted YAML

2. First line does not contain #cloud-config

To verify your YAML, we do have a short script called validate-yaml.py that can validate your user data offline.

Another option is to run the following on an instance to debug userdata provided to the system:

$ cloud-init devel schema --system --annotate

As launching instances in the cloud can cost money and take a bit longer, sometimes it is easier to launch instances locally using Multipass or LXD:

Multipass

Multipass is a cross-platform tool to launch Ubuntu VMs across Linux, Windows, and macOS.

When a user launches a Multipass VM, user data can be passed by adding the –cloud-init flag and the appropriate YAML file containing user data:

$ multipass launch bionic --name test-vm --cloud-init userdata.yaml

Multipass will validate the YAML syntax of the cloud-config file before attempting to start the VM! A nice addition to help save time when experimenting with launching instances with various cloud-configs.

Multipass only supports passing user-data and only as YAML cloud-config files. Passing a script, a MIME archive, or any of the other user-data formats cloud-init supports will result in an error from the YAML syntax validator.

LXD

LXD offers a streamlined user experience for using linux system containers. With LXD, a user can pass:

• user data

• vendor data

• metadata

• network configuration

The following initializes a container with user data:

$ lxc init ubuntu-daily:bionic test-container
$ lxc config set test-container user.user-data - < userdata.yaml
$ lxc start test-container

To avoid the extra commands this can also be done at launch:

$ lxc launch ubuntu-daily:bionic test-container --config=user.user-data="$(cat userdata.yaml)"

Finally, a profile can be setup with the specific data if a user needs to launch this multiple times:

$ lxc profile create dev-user-data
$ lxc profile set dev-user-data user.user-data - < cloud-init-config.yaml
$ lxc launch ubuntu-daily:bionic test-container -p default -p dev-user-data

The above examples all show how to pass user data. To pass other types of configuration data use the config option specified below:

Data	Config Option
user data	user.user-data
vendor data	user.vendor-data
metadata	user.meta-data
network config	user.network-config

See the LXD Instance Configuration docs for more info about configuration values or the LXD Custom Network Configuration document for more about custom network config.

cloud-localds

The cloud-localds command from the cloud-utils package generates a disk with user supplied data. The NoCloud datasouce allows users to provide their own user data, metadata, or network configuration directly to an instance without running a network service. This is helpful for launching local cloud images with QEMU for example.

The following is an example of creating the local disk using the cloud-localds command:

$ cat >user-data <<EOF
#cloud-config
password: password
chpasswd:
  expire: False
ssh_pwauth: True
ssh_authorized_keys:
  - ssh-rsa AAAA...UlIsqdaO+w==
EOF
$ cloud-localds seed.img user-data

The resulting seed.img can then get passed along to a cloud image containing cloud-init. Below is an example of passing the seed.img with QEMU:

$ qemu-system-x86_64 -m 1024 -net nic -net user \
    -hda ubuntu-20.04-server-cloudimg-amd64.img \
    -hdb seed.img

The now booted image will allow for login using the password provided above.

For additional configuration, users can provide much more detailed configuration, including network configuration and metadata:

$ cloud-localds --network-config=network-config-v2.yaml \
  seed.img userdata.yaml metadata.yaml

See the Networking Config Version 2 page for details on the format and config of network configuration. To learn more about the possible values for metadata, check out the NoCloud page.

Where can I learn more?

Below are some videos, blog posts, and white papers about cloud-init from a variety of sources.

• cloud-init - The Good Parts

• cloud-init Summit 2019

• Utilising cloud-init on Microsoft Azure (Whitepaper)

• Cloud Instance Initialization with cloud-init (Whitepaper)

• cloud-init Summit 2018

• cloud-init - The cross-cloud Magic Sauce (PDF)

• cloud-init Summit 2017

• cloud-init - Building clouds one Linux box at a time (Video)

• cloud-init - Building clouds one Linux box at a time (PDF)

• Metadata and cloud-init

• The beauty of cloud-init

• Introduction to cloud-init

Reporting Bugs

The following documents:

1. How to collect information for reporting bugs

2. How to file bugs to the upstream cloud-init project or for distro specific packages

Collect Logs

To aid in debugging, please collect the necessary logs. To do so, run the collect-logs subcommand to produce a tarfile that you can easily upload:

$ sudo cloud-init collect-logs
Wrote /home/ubuntu/cloud-init.tar.gz

If your version of cloud-init does not have the collect-logs subcommand, then please manually collect the base log files by doing the following:

$ sudo dmesg > dmesg.txt
$ sudo journalctl -o short-precise > journal.txt
$ sudo tar -cvf cloud-init.tar dmesg.txt journal.txt /run/cloud-init \
    /var/log/cloud-init.log /var/log/cloud-init-output.log

Report Upstream Bug

Bugs for upstream cloud-init are tracked using Launchpad. To file a bug:

1. Collect the necessary debug logs as described above

2. Create a Launchpad account or login to your existing account

3. Report an upstream cloud-init bug

If debug logs are not provided, you will be asked for them before any further time is spent debugging. If you are unable to obtain the required logs please explain why in the bug.

If your bug is for a specific distro using cloud-init, please first consider reporting it with the upstream distro or confirm that it still occurs with the latest upstream cloud-init code. See below for details on specific distro reporting.

Distro Specific Issues

For issues specific to your distro please use one of the following distro specific reporting mechanisms:

Ubuntu

To report a bug on Ubuntu use the ubuntu-bug command on the affected system to automatically collect the necessary logs and file a bug on Launchpad:

$ ubuntu-bug cloud-init

If that does not work or is not an option, please collect the logs using the commands in the above Collect Logs section and then report the bug on the Ubuntu bug tracker. Make sure to attach your collected logs!

Debian

To file a bug against the Debian package fo cloud-init please use the Debian bug tracker to file against ‘Package: cloud-init’. See the Debian bug reporting wiki wiki page for more details.

Red Hat, CentOS, & Fedora

To file a bug against the Red Hat or Fedora packages of cloud-init please use the Red Hat bugzilla.

SUSE & openSUSE

To file a bug against the SuSE packages of cloud-init please use the SUSE bugzilla.

Arch

To file a bug against the Arch package of cloud-init please use the Arch Linux Bugtracker. See the Arch bug reporting wiki for more details.

User-Data Formats

User data that will be acted upon by cloud-init must be in one of the following types.

Cloud Config Data

Cloud-config is the simplest way to accomplish some things via user-data. Using cloud-config syntax, the user can specify certain things in a human friendly format.

These things include:

• apt upgrade should be run on first boot

• a different apt mirror should be used

• additional apt sources should be added

• certain SSH keys should be imported

• and many more…

Note

This file must be valid YAML syntax.

See the Cloud config examples section for a commented set of examples of supported cloud config formats.

Begins with: #cloud-config or Content-Type: text/cloud-config when using a MIME archive.

Note

New in cloud-init v. 18.4: Cloud config data can also render cloud instance metadata variables using jinja templating. See Instance Metadata for more information.

User-Data Script

Typically used by those who just want to execute a shell script.

Begins with: #! or Content-Type: text/x-shellscript when using a MIME archive.

Note

New in cloud-init v. 18.4: User-data scripts can also render cloud instance metadata variables using jinja templating. See Instance Metadata for more information.

Example

$ cat myscript.sh

#!/bin/sh
echo "Hello World.  The time is now $(date -R)!" | tee /root/output.txt

$ euca-run-instances --key mykey --user-data-file myscript.sh ami-a07d95c9

Kernel Command Line

When using the NoCloud datasource, users can pass user data via the kernel command line parameters. See the NoCloud datasource documentation for more details.

Gzip Compressed Content

Content found to be gzip compressed will be uncompressed. The uncompressed data will then be used as if it were not compressed. This is typically useful because user-data is limited to ~16384 1 bytes.

Mime Multi Part Archive

This list of rules is applied to each part of this multi-part file. Using a mime-multi part file, the user can specify more than one type of data.

For example, both a user data script and a cloud-config type could be specified.

Supported content-types are listed from the cloud-init subcommand make-mime:

$ cloud-init devel make-mime --list-types
cloud-boothook
cloud-config
cloud-config-archive
cloud-config-jsonp
jinja2
part-handler
upstart-job
x-include-once-url
x-include-url
x-shellscript
x-shellscript-per-boot
x-shellscript-per-instance
x-shellscript-per-once

Helper subcommand to generate mime messages

The cloud-init subcommand can generate MIME multi-part files: make-mime.

make-mime subcommand takes pairs of (filename, “text/” mime subtype) separated by a colon (e.g. config.yaml:cloud-config) and emits a MIME multipart message to stdout.

Examples

Create userdata containing both a cloud-config (config.yaml) and a shell script (script.sh)

$ cloud-init devel make-mime -a config.yaml:cloud-config -a script.sh:x-shellscript > userdata

Create userdata containing 3 shell scripts:

• always.sh - Run every boot

• instance.sh - Run once per instance

• once.sh - Run once

$ cloud-init devel make-mime -a always.sh:x-shellscript-per-boot -a instance.sh:x-shellscript-per-instance -a once.sh:x-shellscript-per-once

Include File

This content is a include file.

The file contains a list of urls, one per line. Each of the URLs will be read, and their content will be passed through this same set of rules. I.e., the content read from the URL can be gzipped, mime-multi-part, or plain text. If an error occurs reading a file the remaining files will not be read.

Begins with: #include or Content-Type: text/x-include-url when using a MIME archive.

Cloud Boothook

This content is boothook data. It is stored in a file under /var/lib/cloud and then executed immediately. This is the earliest hook available. Note, that there is no mechanism provided for running only once. The boothook must take care of this itself.

It is provided with the instance id in the environment variable INSTANCE_ID. This could be made use of to provide a ‘once-per-instance’ type of functionality.

Begins with: #cloud-boothook or Content-Type: text/cloud-boothook when using a MIME archive.

Part Handler

This is a part-handler: It contains custom code for either supporting new mime-types in multi-part user data, or overriding the existing handlers for supported mime-types. It will be written to a file in /var/lib/cloud/data based on its filename (which is generated).

This must be python code that contains a list_types function and a handle_part function. Once the section is read the list_types method will be called. It must return a list of mime-types that this part-handler handles. Because mime parts are processed in order, a part-handler part must precede any parts with mime-types it is expected to handle in the same user data.

The handle_part function must be defined like:

def handle_part(data, ctype, filename, payload):
  # data = the cloudinit object
  # ctype = "__begin__", "__end__", or the mime-type of the part that is being handled.
  # filename = the filename of the part (or a generated filename if none is present in mime data)
  # payload = the parts' content

Cloud-init will then call the handle_part function once before it handles any parts, once per part received, and once after all parts have been handled. The '__begin__' and '__end__' sentinels allow the part handler to do initialization or teardown before or after receiving any parts.

Begins with: #part-handler or Content-Type: text/part-handler when using a MIME archive.

Example

#part-handler

def list_types():
    # return a list of mime-types that are handled by this module
    return(["text/plain", "text/go-cubs-go"])

def handle_part(data, ctype, filename, payload):
    # data: the cloudinit object
    # ctype: '__begin__', '__end__', or the specific mime-type of the part
    # filename: the filename for the part, or dynamically generated part if
    #           no filename is given attribute is present
    # payload: the content of the part (empty for begin or end)
    if ctype == "__begin__":
       print("my handler is beginning")
       return
    if ctype == "__end__":
       print("my handler is ending")
       return

    print(f"==== received ctype={ctype} filename={filename} ====")
    print(payload)
    print(f"==== end ctype={ctype} filename={filename}")

Also this blog post offers another example for more advanced usage.

Upstart Job

Content is placed into a file in /etc/init, and will be consumed by upstart as any other upstart job.

Begins with: #upstart-job or Content-Type: text/upstart-job when using a MIME archive.

Disabling User-Data

Cloud-init can be configured to ignore any user-data provided to instance. This allows custom images to prevent users from accidentally breaking closed appliances. Setting allow_userdata: false in the configuration will disable cloud-init from processing user-data.

1

See your cloud provider for applicable user-data size limitations…

Cloud config examples

Including users and groups

#cloud-config
# Add groups to the system
# The following example adds the ubuntu group with members 'root' and 'sys'
# and the empty group cloud-users.
groups:
  - ubuntu: [root,sys]
  - cloud-users

# Add users to the system. Users are added after groups are added.
# Note: Most of these configuration options will not be honored if the user
#       already exists. Following options are the exceptions and they are
#       applicable on already-existing users:
#       - 'plain_text_passwd', 'hashed_passwd', 'lock_passwd', 'sudo',
#         'ssh_authorized_keys', 'ssh_redirect_user'.
users:
  - default
  - name: foobar
    gecos: Foo B. Bar
    primary_group: foobar
    groups: users
    selinux_user: staff_u
    expiredate: '2032-09-01'
    ssh_import_id:
      - lp:falcojr
      - gh:TheRealFalcon
    lock_passwd: false
    passwd: $6$j212wezy$7H/1LT4f9/N3wpgNunhsIqtMj62OKiS3nyNwuizouQc3u7MbYCarYeAHWYPYb2FT.lbioDm2RrkJPb9BZMN1O/
  - name: barfoo
    gecos: Bar B. Foo
    sudo: ALL=(ALL) NOPASSWD:ALL
    groups: users, admin
    ssh_import_id:
      - lp:falcojr
      - gh:TheRealFalcon
    lock_passwd: true
    ssh_authorized_keys:
      - <ssh pub key 1>
      - <ssh pub key 2>
  - name: cloudy
    gecos: Magic Cloud App Daemon User
    inactive: '5'
    system: true
  - name: fizzbuzz
    sudo: False
    ssh_authorized_keys:
      - <ssh pub key 1>
      - <ssh pub key 2>
  - snapuser: joe@joeuser.io
  - name: nosshlogins
    ssh_redirect_user: true

# Valid Values:
#   name: The user's login name
#   expiredate: Date on which the user's account will be disabled.
#   gecos: The user name's real name, i.e. "Bob B. Smith"
#   homedir: Optional. Set to the local path you want to use. Defaults to
#           /home/<username>
#   primary_group: define the primary group. Defaults to a new group created
#           named after the user.
#   groups:  Optional. Additional groups to add the user to. Defaults to none
#   selinux_user:  Optional. The SELinux user for the user's login, such as
#           "staff_u". When this is omitted the system will select the default
#           SELinux user.
#   lock_passwd: Defaults to true. Lock the password to disable password login
#   inactive: Number of days after password expires until account is disabled
#   passwd: The hash -- not the password itself -- of the password you want
#           to use for this user. You can generate a hash via:
#               mkpasswd --method=SHA-512 --rounds=4096
#           (the above command would create from stdin an SHA-512 password hash
#           with 4096 salt rounds)
#
#           Please note: while the use of a hashed password is better than
#               plain text, the use of this feature is not ideal. Also,
#               using a high number of salting rounds will help, but it should
#               not be relied upon.
#
#               To highlight this risk, running John the Ripper against the
#               example hash above, with a readily available wordlist, revealed
#               the true password in 12 seconds on a i7-2620QM.
#
#               In other words, this feature is a potential security risk and is
#               provided for your convenience only. If you do not fully trust the
#               medium over which your cloud-config will be transmitted, then you
#               should not use this feature.
#
#   no_create_home: When set to true, do not create home directory.
#   no_user_group: When set to true, do not create a group named after the user.
#   no_log_init: When set to true, do not initialize lastlog and faillog database.
#   ssh_import_id: Optional. Import SSH ids
#   ssh_authorized_keys: Optional. [list] Add keys to user's authorized keys file
#   ssh_redirect_user: Optional. [bool] Set true to block ssh logins for cloud
#       ssh public keys and emit a message redirecting logins to
#       use <default_username> instead. This option only disables cloud
#       provided public-keys. An error will be raised if ssh_authorized_keys
#       or ssh_import_id is provided for the same user.
#
#   sudo: Defaults to none. Accepts a sudo rule string, a list of sudo rule
#         strings or False to explicitly deny sudo usage. Examples:
#
#         Allow a user unrestricted sudo access.
#             sudo:  ALL=(ALL) NOPASSWD:ALL
#
#         Adding multiple sudo rule strings.
#             sudo:
#               - ALL=(ALL) NOPASSWD:/bin/mysql
#               - ALL=(ALL) ALL
#
#         Prevent sudo access for a user.
#             sudo: False
#
#         Note: Please double check your syntax and make sure it is valid.
#               cloud-init does not parse/check the syntax of the sudo
#               directive.
#   system: Create the user as a system user. This means no home directory.
#   snapuser: Create a Snappy (Ubuntu-Core) user via the snap create-user
#             command available on Ubuntu systems.  If the user has an account
#             on the Ubuntu SSO, specifying the email will allow snap to
#             request a username and any public ssh keys and will import
#             these into the system with username specified by SSO account.
#             If 'username' is not set in SSO, then username will be the
#             shortname before the email domain.
#

# Default user creation:
#
# Unless you define users, you will get a 'ubuntu' user on Ubuntu systems with the
# legacy permission (no password sudo, locked user, etc). If however, you want
# to have the 'ubuntu' user in addition to other users, you need to instruct
# cloud-init that you also want the default user. To do this use the following
# syntax:
#   users:
#     - default
#     - bob
#     - ....
#  foobar: ...
#
# users[0] (the first user in users) overrides the user directive.
#
# The 'default' user above references the distro's config:
# system_info:
#   default_user:
#     name: Ubuntu
#     plain_text_passwd: 'ubuntu'
#     home: /home/ubuntu
#     shell: /bin/bash
#     lock_passwd: True
#     gecos: Ubuntu
#     groups: [adm, audio, cdrom, dialout, floppy, video, plugdev, dip, netdev]

Writing out arbitrary files

#cloud-config
# vim: syntax=yaml
#
# This is the configuration syntax that the write_files module
# will know how to understand. Encoding can be given b64 or gzip or (gz+b64).
# The content will be decoded accordingly and then written to the path that is
# provided. 
#
# Note: Content strings here are truncated for example purposes.
write_files:
- encoding: b64
  content: CiMgVGhpcyBmaWxlIGNvbnRyb2xzIHRoZSBzdGF0ZSBvZiBTRUxpbnV4...
  owner: root:root
  path: /etc/sysconfig/selinux
  permissions: '0644'
- content: |
    # My new /etc/sysconfig/samba file

    SMBDOPTIONS="-D"
  path: /etc/sysconfig/samba
- content: !!binary |
    f0VMRgIBAQAAAAAAAAAAAAIAPgABAAAAwARAAAAAAABAAAAAAAAAAJAVAAAAAAAAAAAAAEAAOAAI
    AEAAHgAdAAYAAAAFAAAAQAAAAAAAAABAAEAAAAAAAEAAQAAAAAAAwAEAAAAAAADAAQAAAAAAAAgA
    AAAAAAAAAwAAAAQAAAAAAgAAAAAAAAACQAAAAAAAAAJAAAAAAAAcAAAAAAAAABwAAAAAAAAAAQAA
    ....
  path: /bin/arch
  permissions: '0555'
- encoding: gzip
  content: !!binary |
    H4sIAIDb/U8C/1NW1E/KzNMvzuBKTc7IV8hIzcnJVyjPL8pJ4QIA6N+MVxsAAAA=
  path: /usr/bin/hello
  permissions: '0755'

Adding a yum repository

#cloud-config
# vim: syntax=yaml
#
# Add yum repository configuration to the system
#
# The following example adds the file /etc/yum.repos.d/epel_testing.repo
# which can then subsequently be used by yum for later operations.
yum_repos:
  # The name of the repository
  epel-testing:
    # Any repository configuration options
    # See: man yum.conf
    #
    # This one is required!
    baseurl: http://download.fedoraproject.org/pub/epel/testing/5/$basearch
    enabled: false
    failovermethod: priority
    gpgcheck: true
    gpgkey: file:///etc/pki/rpm-gpg/RPM-GPG-KEY-EPEL
    name: Extra Packages for Enterprise Linux 5 - Testing

Configure an instances trusted CA certificates

#cloud-config
#
# This is an example file to configure an instance's trusted CA certificates
# system-wide for SSL/TLS trust establishment when the instance boots for the
# first time.
#
# Make sure that this file is valid yaml before starting instances.
# It should be passed as user-data when starting the instance.

ca_certs:
  # If present and set to True, the 'remove_defaults' parameter will remove
  # all the default trusted CA certificates that are normally shipped with
  # Ubuntu.
  # This is mainly for paranoid admins - most users will not need this
  # functionality.
  remove_defaults: true

  # If present, the 'trusted' parameter should contain a certificate (or list
  # of certificates) to add to the system as trusted CA certificates.
  # Pay close attention to the YAML multiline list syntax.  The example shown
  # here is for a list of multiline certificates.
  trusted: 
  - |
   -----BEGIN CERTIFICATE-----
   YOUR-ORGS-TRUSTED-CA-CERT-HERE
   -----END CERTIFICATE-----
  - |
   -----BEGIN CERTIFICATE-----
   YOUR-ORGS-TRUSTED-CA-CERT-HERE
   -----END CERTIFICATE-----

Install and run chef recipes

#cloud-config
#
# This is an example file to automatically install chef-client and run a
# list of recipes when the instance boots for the first time.
# Make sure that this file is valid yaml before starting instances.
# It should be passed as user-data when starting the instance.

# The default is to install from packages.

# Key from https://packages.chef.io/chef.asc
apt:
  sources:
    source1:
      source: "deb http://packages.chef.io/repos/apt/stable $RELEASE main"
      key: |
        -----BEGIN PGP PUBLIC KEY BLOCK-----
        Version: GnuPG v1.4.12 (Darwin)
        Comment: GPGTools - http://gpgtools.org

        mQGiBEppC7QRBADfsOkZU6KZK+YmKw4wev5mjKJEkVGlus+NxW8wItX5sGa6kdUu
        twAyj7Yr92rF+ICFEP3gGU6+lGo0Nve7KxkN/1W7/m3G4zuk+ccIKmjp8KS3qn99
        dxy64vcji9jIllVa+XXOGIp0G8GEaj7mbkixL/bMeGfdMlv8Gf2XPpp9vwCgn/GC
        JKacfnw7MpLKUHOYSlb//JsEAJqao3ViNfav83jJKEkD8cf59Y8xKia5OpZqTK5W
        ShVnNWS3U5IVQk10ZDH97Qn/YrK387H4CyhLE9mxPXs/ul18ioiaars/q2MEKU2I
        XKfV21eMLO9LYd6Ny/Kqj8o5WQK2J6+NAhSwvthZcIEphcFignIuobP+B5wNFQpe
        DbKfA/0WvN2OwFeWRcmmd3Hz7nHTpcnSF+4QX6yHRF/5BgxkG6IqBIACQbzPn6Hm
        sMtm/SVf11izmDqSsQptCrOZILfLX/mE+YOl+CwWSHhl+YsFts1WOuh1EhQD26aO
        Z84HuHV5HFRWjDLw9LriltBVQcXbpfSrRP5bdr7Wh8vhqJTPjrQnT3BzY29kZSBQ
        YWNrYWdlcyA8cGFja2FnZXNAb3BzY29kZS5jb20+iGAEExECACAFAkppC7QCGwMG
        CwkIBwMCBBUCCAMEFgIDAQIeAQIXgAAKCRApQKupg++Caj8sAKCOXmdG36gWji/K
        +o+XtBfvdMnFYQCfTCEWxRy2BnzLoBBFCjDSK6sJqCu0IENIRUYgUGFja2FnZXMg
        PHBhY2thZ2VzQGNoZWYuaW8+iGIEExECACIFAlQwYFECGwMGCwkIBwMCBhUIAgkK
        CwQWAgMBAh4BAheAAAoJEClAq6mD74JqX94An26z99XOHWpLN8ahzm7cp13t4Xid
        AJ9wVcgoUBzvgg91lKfv/34cmemZn7kCDQRKaQu0EAgAg7ZLCVGVTmLqBM6njZEd
        Zbv+mZbvwLBSomdiqddE6u3eH0X3GuwaQfQWHUVG2yedyDMiG+EMtCdEeeRebTCz
        SNXQ8Xvi22hRPoEsBSwWLZI8/XNg0n0f1+GEr+mOKO0BxDB2DG7DA0nnEISxwFkK
        OFJFebR3fRsrWjj0KjDxkhse2ddU/jVz1BY7Nf8toZmwpBmdozETMOTx3LJy1HZ/
        Te9FJXJMUaB2lRyluv15MVWCKQJro4MQG/7QGcIfrIZNfAGJ32DDSjV7/YO+IpRY
        IL4CUBQ65suY4gYUG4jhRH6u7H1p99sdwsg5OIpBe/v2Vbc/tbwAB+eJJAp89Zeu
        twADBQf/ZcGoPhTGFuzbkcNRSIz+boaeWPoSxK2DyfScyCAuG41CY9+g0HIw9Sq8
        DuxQvJ+vrEJjNvNE3EAEdKl/zkXMZDb1EXjGwDi845TxEMhhD1dDw2qpHqnJ2mtE
        WpZ7juGwA3sGhi6FapO04tIGacCfNNHmlRGipyq5ZiKIRq9mLEndlECr8cwaKgkS
        0wWu+xmMZe7N5/t/TK19HXNh4tVacv0F3fYK54GUjt2FjCQV75USnmNY4KPTYLXA
        dzC364hEMlXpN21siIFgB04w+TXn5UF3B4FfAy5hevvr4DtV4MvMiGLu0oWjpaLC
        MpmrR3Ny2wkmO0h+vgri9uIP06ODWIhJBBgRAgAJBQJKaQu0AhsMAAoJEClAq6mD
        74Jq4hIAoJ5KrYS8kCwj26SAGzglwggpvt3CAJ0bekyky56vNqoegB+y4PQVDv4K
        zA==
        =IxPr
        -----END PGP PUBLIC KEY BLOCK-----

chef:

  # Valid values are 'accept' and 'accept-no-persist'
  chef_license: "accept"

  # Valid values are 'gems' and 'packages' and 'omnibus'
  install_type: "packages"

  # Boolean: run 'install_type' code even if chef-client
  #          appears already installed.
  force_install: false

  # Chef settings
  server_url: "https://chef.yourorg.com"

  # Node Name
  # Defaults to the instance-id if not present
  node_name: "your-node-name"

  # Environment
  # Defaults to '_default' if not present
  environment: "production"

  # Default validation name is chef-validator
  validation_name: "yourorg-validator"
  # if validation_cert's value is "system" then it is expected
  # that the file already exists on the system.
  validation_cert: |
    -----BEGIN RSA PRIVATE KEY-----
    YOUR-ORGS-VALIDATION-KEY-HERE
    -----END RSA PRIVATE KEY-----

  # A run list for a first boot json, an example (not required)
  run_list:
    - "recipe[apache2]"
    - "role[db]"

  # Specify a list of initial attributes used by the cookbooks
  initial_attributes:
    apache:
      prefork:
        maxclients: 100
      keepalive: "off"

  # if install_type is 'omnibus', change the url to download
  omnibus_url: "https://www.chef.io/chef/install.sh"

  # if install_type is 'omnibus', pass pinned version string
  # to the install script
  omnibus_version: "12.3.0"

  # If encrypted data bags are used, the client needs to have a secrets file
  # configured to decrypt them
  encrypted_data_bag_secret: "/etc/chef/encrypted_data_bag_secret"

# Capture all subprocess output into a logfile
# Useful for troubleshooting cloud-init issues
output: {all: '| tee -a /var/log/cloud-init-output.log'}

Add primary apt repositories

#cloud-config

# Add primary apt repositories
#
# To add 3rd party repositories, see cloud-config-apt.txt or the
# Additional apt configuration and repositories section.
#
#
# Default: auto select based on cloud metadata
#  in ec2, the default is <region>.archive.ubuntu.com
# apt:
#   primary:
#     - arches [default]
#       uri:
#     use the provided mirror
#       search:
#     search the list for the first mirror.
#     this is currently very limited, only verifying that
#     the mirror is dns resolvable or an IP address
#
# if neither mirror is set (the default)
# then use the mirror provided by the DataSource found.
# In EC2, that means using <region>.ec2.archive.ubuntu.com
#
# if no mirror is provided by the DataSource, but 'search_dns' is
# true, then search for dns names '<distro>-mirror' in each of
# - fqdn of this host per cloud metadata
# - localdomain
# - no domain (which would search domains listed in /etc/resolv.conf)
# If there is a dns entry for <distro>-mirror, then it is assumed that there
# is a distro mirror at http://<distro>-mirror.<domain>/<distro>
#
# That gives the cloud provider the opportunity to set mirrors of a distro
# up and expose them only by creating dns entries.
#
# if none of that is found, then the default distro mirror is used
apt:
  primary:
    - arches: [default]
      uri: http://us.archive.ubuntu.com/ubuntu/
# or
apt:
  primary:
    - arches: [default]
      search:
        - http://local-mirror.mydomain
        - http://archive.ubuntu.com
# or
apt:
  primary:
    - arches: [default]
      search_dns: True

Run commands on first boot

#cloud-config

# boot commands
# default: none
# this is very similar to runcmd, but commands run very early
# in the boot process, only slightly after a 'boothook' would run.
# bootcmd should really only be used for things that could not be
# done later in the boot process.  bootcmd is very much like
# boothook, but possibly with more friendly.
# - bootcmd will run on every boot
# - the INSTANCE_ID variable will be set to the current instance id.
# - you can use 'cloud-init-per' command to help only run once
bootcmd:
  - echo 192.168.1.130 us.archive.ubuntu.com >> /etc/hosts
  - [ cloud-init-per, once, mymkfs, mkfs, /dev/vdb ]

#cloud-config

# run commands
# default: none
# runcmd contains a list of either lists or a string
# each item will be executed in order at rc.local like level with
# output to the console
# - runcmd only runs during the first boot
# - if the item is a list, the items will be properly executed as if
#   passed to execve(3) (with the first arg as the command).
# - if the item is a string, it will be simply written to the file and
#   will be interpreted by 'sh'
#
# Note, that the list has to be proper yaml, so you have to quote
# any characters yaml would eat (':' can be problematic)
runcmd:
 - [ ls, -l, / ]
 - [ sh, -xc, "echo $(date) ': hello world!'" ]
 - [ sh, -c, echo "=========hello world'=========" ]
 - ls -l /root
 # Note: Don't write files to /tmp from cloud-init use /run/somedir instead.
 # Early boot environments can race systemd-tmpfiles-clean LP: #1707222.
 - mkdir /run/mydir
 - [ wget, "http://slashdot.org", -O, /run/mydir/index.html ]

Alter the completion message

#cloud-config

# final_message
# default: cloud-init boot finished at $TIMESTAMP. Up $UPTIME seconds
# this message is written by cloud-final when the system is finished
# its first boot
final_message: "The system is finally up, after $UPTIME seconds"

Install arbitrary packages

#cloud-config

# Install additional packages on first boot
#
# Default: none
#
# if packages are specified, this package_update will be set to true
#
# packages may be supplied as a single package name or as a list
# with the format [<package>, <version>] wherein the specific
# package version will be installed.
packages:
 - pwgen
 - pastebinit
 - [libpython2.7, 2.7.3-0ubuntu3.1]

Update apt database on first boot

#cloud-config
# Update apt database on first boot (run 'apt-get update').
# Note, if packages are given, or package_upgrade is true, then
# update will be done independent of this setting.
#
# Default: false
package_update: true

Run apt or yum upgrade

#cloud-config

# Upgrade the instance on first boot
#
# Default: false
package_upgrade: true

Adjust mount points mounted

#cloud-config

# set up mount points
# 'mounts' contains a list of lists
#  the inner list are entries for an /etc/fstab line
#  ie : [ fs_spec, fs_file, fs_vfstype, fs_mntops, fs-freq, fs_passno ]
#
# default:
# mounts:
#  - [ ephemeral0, /mnt ]
#  - [ swap, none, swap, sw, 0, 0 ]
#
# in order to remove a previously listed mount (ie, one from defaults)
# list only the fs_spec.  For example, to override the default, of
# mounting swap:
# - [ swap ]
# or
# - [ swap, null ]
#
# - if a device does not exist at the time, an entry will still be
#   written to /etc/fstab.
# - '/dev' can be omitted for device names that begin with: xvd, sd, hd, vd
# - if an entry does not have all 6 fields, they will be filled in
#   with values from 'mount_default_fields' below.
#
# Note, that you should set 'nofail' (see man fstab) for volumes that may not
# be attached at instance boot (or reboot).
#
mounts:
 - [ ephemeral0, /mnt, auto, "defaults,noexec" ]
 - [ sdc, /opt/data ]
 - [ xvdh, /opt/data, "auto", "defaults,nofail", "0", "0" ]
 - [ dd, /dev/zero ]

# mount_default_fields
# These values are used to fill in any entries in 'mounts' that are not
# complete.  This must be an array, and must have 6 fields.
mount_default_fields: [ None, None, "auto", "defaults,nofail", "0", "2" ]


# swap can also be set up by the 'mounts' module
# default is to not create any swap files, because 'size' is set to 0
swap:
  filename: /swap.img
  size: "auto" # or size in bytes
  maxsize: size in bytes

Configure instances SSH keys

#cloud-config

# add each entry to ~/.ssh/authorized_keys for the configured user or the
# first user defined in the user definition directive.
ssh_authorized_keys:
  - ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAGEA3FSyQwBI6Z+nCSjUUk8EEAnnkhXlukKoUPND/RRClWz2s5TCzIkd3Ou5+Cyz71X0XmazM3l5WgeErvtIwQMyT1KjNoMhoJMrJnWqQPOt5Q8zWd9qG7PBl9+eiH5qV7NZ mykey@host
  - ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAQEA3I7VUf2l5gSn5uavROsc5HRDpZdQueUq5ozemNSj8T7enqKHOEaFoU2VoPgGEWC9RyzSQVeyD6s7APMcE82EtmW4skVEgEGSbDc1pvxzxtchBj78hJP6Cf5TCMFSXw+Fz5rF1dR23QDbN1mkHs7adr8GW4kSWqU7Q7NDwfIrJJtO7Hi42GyXtvEONHbiRPOe8stqUly7MvUoN+5kfjBM8Qqpfl2+FNhTYWpMfYdPUnE7u536WqzFmsaqJctz3gBxH9Ex7dFtrxR4qiqEr9Qtlu3xGn7Bw07/+i1D+ey3ONkZLN+LQ714cgj8fRS4Hj29SCmXp5Kt5/82cD/VN3NtHw== smoser@brickies

# Send pre-generated SSH private keys to the server
# If these are present, they will be written to /etc/ssh and
# new random keys will not be generated
#  in addition to 'rsa' and 'dsa' as shown below, 'ecdsa' is also supported
ssh_keys:
  rsa_private: |
    -----BEGIN RSA PRIVATE KEY-----
    MIIBxwIBAAJhAKD0YSHy73nUgysO13XsJmd4fHiFyQ+00R7VVu2iV9Qcon2LZS/x
    1cydPZ4pQpfjEha6WxZ6o8ci/Ea/w0n+0HGPwaxlEG2Z9inNtj3pgFrYcRztfECb
    1j6HCibZbAzYtwIBIwJgO8h72WjcmvcpZ8OvHSvTwAguO2TkR6mPgHsgSaKy6GJo
    PUJnaZRWuba/HX0KGyhz19nPzLpzG5f0fYahlMJAyc13FV7K6kMBPXTRR6FxgHEg
    L0MPC7cdqAwOVNcPY6A7AjEA1bNaIjOzFN2sfZX0j7OMhQuc4zP7r80zaGc5oy6W
    p58hRAncFKEvnEq2CeL3vtuZAjEAwNBHpbNsBYTRPCHM7rZuG/iBtwp8Rxhc9I5w
    ixvzMgi+HpGLWzUIBS+P/XhekIjPAjA285rVmEP+DR255Ls65QbgYhJmTzIXQ2T9
    luLvcmFBC6l35Uc4gTgg4ALsmXLn71MCMGMpSWspEvuGInayTCL+vEjmNBT+FAdO
    W7D4zCpI43jRS9U06JVOeSc9CDk2lwiA3wIwCTB/6uc8Cq85D9YqpM10FuHjKpnP
    REPPOyrAspdeOAV+6VKRavstea7+2DZmSUgE
    -----END RSA PRIVATE KEY-----

  rsa_public: ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAGEAoPRhIfLvedSDKw7XdewmZ3h8eIXJD7TRHtVW7aJX1ByifYtlL/HVzJ09nilCl+MSFrpbFnqjxyL8Rr/DSf7QcY/BrGUQbZn2Kc22PemAWthxHO18QJvWPocKJtlsDNi3 smoser@localhost

  dsa_private: |
    -----BEGIN DSA PRIVATE KEY-----
    MIIBuwIBAAKBgQDP2HLu7pTExL89USyM0264RCyWX/CMLmukxX0Jdbm29ax8FBJT
    pLrO8TIXVY5rPAJm1dTHnpuyJhOvU9G7M8tPUABtzSJh4GVSHlwaCfycwcpLv9TX
    DgWIpSj+6EiHCyaRlB1/CBp9RiaB+10QcFbm+lapuET+/Au6vSDp9IRtlQIVAIMR
    8KucvUYbOEI+yv+5LW9u3z/BAoGBAI0q6JP+JvJmwZFaeCMMVxXUbqiSko/P1lsa
    LNNBHZ5/8MOUIm8rB2FC6ziidfueJpqTMqeQmSAlEBCwnwreUnGfRrKoJpyPNENY
    d15MG6N5J+z81sEcHFeprryZ+D3Ge9VjPq3Tf3NhKKwCDQ0240aPezbnjPeFm4mH
    bYxxcZ9GAoGAXmLIFSQgiAPu459rCKxT46tHJtM0QfnNiEnQLbFluefZ/yiI4DI3
    8UzTCOXLhUA7ybmZha+D/csj15Y9/BNFuO7unzVhikCQV9DTeXX46pG4s1o23JKC
    /QaYWNMZ7kTRv+wWow9MhGiVdML4ZN4XnifuO5krqAybngIy66PMEoQCFEIsKKWv
    99iziAH0KBMVbxy03Trz
    -----END DSA PRIVATE KEY-----

  dsa_public: ssh-dss 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 smoser@localhost

# By default, the fingerprints of the authorized keys for the users
# cloud-init adds are printed to the console. Setting
# no_ssh_fingerprints to true suppresses this output.
no_ssh_fingerprints: false

# By default, (most) ssh host keys are printed to the console. Setting
# emit_keys_to_console to false suppresses this output.
ssh:
  emit_keys_to_console: false

Additional apt configuration and repositories

#cloud-config
# apt_pipelining (configure Acquire::http::Pipeline-Depth)
# Default: disables HTTP pipelining. Certain web servers, such
# as S3 do not pipeline properly (LP: #948461).
# Valid options:
#   False/default: Disables pipelining for APT
#   None/Unchanged: Use OS default
#   Number: Set pipelining to some number (not recommended)
apt_pipelining: False

## apt config via system_info:
# under the 'system_info', you can customize cloud-init's interaction
# with apt.
#  system_info:
#    apt_get_command: [command, argument, argument]
#    apt_get_upgrade_subcommand: dist-upgrade
#
# apt_get_command:
#  To specify a different 'apt-get' command, set 'apt_get_command'.
#  This must be a list, and the subcommand (update, upgrade) is appended to it.
#  default is:
#    ['apt-get', '--option=Dpkg::Options::=--force-confold',
#     '--option=Dpkg::options::=--force-unsafe-io', '--assume-yes', '--quiet']
#
# apt_get_upgrade_subcommand: "dist-upgrade"
#  Specify a different subcommand for 'upgrade. The default is 'dist-upgrade'.
#  This is the subcommand that is invoked for package_upgrade.
#
# apt_get_wrapper:
#   command: eatmydata
#   enabled: [True, False, "auto"]
#

# Install additional packages on first boot
#
# Default: none
#
# if packages are specified, this package_update will be set to true

packages: ['pastebinit']

apt:
  # The apt config consists of two major "areas".
  #
  # On one hand there is the global configuration for the apt feature.
  #
  # On one hand (down in this file) there is the source dictionary which allows
  # to define various entries to be considered by apt.

  ##############################################################################
  # Section 1: global apt configuration
  #
  # The following examples number the top keys to ease identification in
  # discussions.

  # 1.1 preserve_sources_list
  #
  # Preserves the existing /etc/apt/sources.list
  # Default: false - do overwrite sources_list. If set to true then any
  # "mirrors" configuration will have no effect.
  # Set to true to avoid affecting sources.list. In that case only
  # "extra" source specifications will be written into
  # /etc/apt/sources.list.d/*
  preserve_sources_list: true

  # 1.2 disable_suites
  #
  # This is an empty list by default, so nothing is disabled.
  #
  # If given, those suites are removed from sources.list after all other
  # modifications have been made.
  # Suites are even disabled if no other modification was made,
  # but not if is preserve_sources_list is active.
  # There is a special alias "$RELEASE" as in the sources that will be replace
  # by the matching release.
  #
  # To ease configuration and improve readability the following common ubuntu
  # suites will be automatically mapped to their full definition.
  # updates   => $RELEASE-updates
  # backports => $RELEASE-backports
  # security  => $RELEASE-security
  # proposed  => $RELEASE-proposed
  # release   => $RELEASE
  #
  # There is no harm in specifying a suite to be disabled that is not found in
  # the source.list file (just a no-op then)
  #
  # Note: Lines don't get deleted, but disabled by being converted to a comment.
  # The following example disables all usual defaults except $RELEASE-security.
  # On top it disables a custom suite called "mysuite"
  disable_suites: [$RELEASE-updates, backports, $RELEASE, mysuite]

  # 1.3 primary/security archives
  #
  # Default: none - instead it is auto select based on cloud metadata
  # so if neither "uri" nor "search", nor "search_dns" is set (the default)
  # then use the mirror provided by the DataSource found.
  # In EC2, that means using <region>.ec2.archive.ubuntu.com
  #
  # define a custom (e.g. localized) mirror that will be used in sources.list
  # and any custom sources entries for deb / deb-src lines.
  #
  # One can set primary and security mirror to different uri's
  # the child elements to the keys primary and secondary are equivalent
  primary:
    # arches is list of architectures the following config applies to
    # the special keyword "default" applies to any architecture not explicitly
    # listed.
    - arches: [amd64, i386, default]
      # uri is just defining the target as-is
      uri: http://us.archive.ubuntu.com/ubuntu
      #
      # via search one can define lists that are tried one by one.
      # The first with a working DNS resolution (or if it is an IP) will be
      # picked. That way one can keep one configuration for multiple
      # subenvironments that select the working one.
      search:
        - http://cool.but-sometimes-unreachable.com/ubuntu
        - http://us.archive.ubuntu.com/ubuntu
      # if no mirror is provided by uri or search but 'search_dns' is
      # true, then search for dns names '<distro>-mirror' in each of
      # - fqdn of this host per cloud metadata
      # - localdomain
      # - no domain (which would search domains listed in /etc/resolv.conf)
      # If there is a dns entry for <distro>-mirror, then it is assumed that
      # there is a distro mirror at http://<distro>-mirror.<domain>/<distro>
      #
      # That gives the cloud provider the opportunity to set mirrors of a distro
      # up and expose them only by creating dns entries.
      #
      # if none of that is found, then the default distro mirror is used
      search_dns: true
      #
      # If multiple of a category are given
      #   1. uri
      #   2. search
      #   3. search_dns
      # the first defining a valid mirror wins (in the order as defined here,
      # not the order as listed in the config).
      #
      # Additionally, if the repository requires a custom signing key, it can be
      # specified via the same fields as for custom sources:
      #   'keyid': providing a key to import via shortid or fingerprint
      #   'key': providing a raw PGP key
      #   'keyserver': specify an alternate keyserver to pull keys from that
      #                were specified by keyid
    - arches: [s390x, arm64]
      # as above, allowing to have one config for different per arch mirrors
  # security is optional, if not defined it is set to the same value as primary
  security:
    - uri: http://security.ubuntu.com/ubuntu
      arches: [default]
  # If search_dns is set for security the searched pattern is:
  #   <distro>-security-mirror

  # if no mirrors are specified at all, or all lookups fail it will try
  # to get them from the cloud datasource and if those neither provide one fall
  # back to:
  #   primary: http://archive.ubuntu.com/ubuntu
  #   security: http://security.ubuntu.com/ubuntu

  # 1.4 sources_list
  #
  # Provide a custom template for rendering sources.list
  # without one provided cloud-init uses builtin templates for
  # ubuntu and debian.
  # Within these sources.list templates you can use the following replacement
  # variables (all have sane Ubuntu defaults, but mirrors can be overwritten
  # as needed (see above)):
  # => $RELEASE, $MIRROR, $PRIMARY, $SECURITY
  sources_list: | # written by cloud-init custom template
    deb $MIRROR $RELEASE main restricted
    deb-src $MIRROR $RELEASE main restricted
    deb $PRIMARY $RELEASE universe restricted
    deb $SECURITY $RELEASE-security multiverse

  # 1.5 conf
  #
  # Any apt config string that will be made available to apt
  # see the APT.CONF(5) man page for details what can be specified
  conf: | # APT config
    APT {
      Get {
        Assume-Yes "true";
        Fix-Broken "true";
      };
    };

  # 1.6 (http_|ftp_|https_)proxy
  #
  # Proxies are the most common apt.conf option, so that for simplified use
  # there is a shortcut for those. Those get automatically translated into the
  # correct Acquire::*::Proxy statements.
  #
  # note: proxy actually being a short synonym to http_proxy
  proxy: http://[[user][:pass]@]host[:port]/
  http_proxy: http://[[user][:pass]@]host[:port]/
  ftp_proxy: ftp://[[user][:pass]@]host[:port]/
  https_proxy: https://[[user][:pass]@]host[:port]/

  # 1.7 add_apt_repo_match
  #
  # 'source' entries in apt-sources that match this python regex
  # expression will be passed to add-apt-repository
  # The following example is also the builtin default if nothing is specified
  add_apt_repo_match: '^[\w-]+:\w'


  ##############################################################################
  # Section 2: source list entries
  #
  # This is a dictionary (unlike most block/net which are lists)
  #
  # The key of each source entry is the filename and will be prepended by
  # /etc/apt/sources.list.d/ if it doesn't start with a '/'.
  # If it doesn't end with .list it will be appended so that apt picks up its
  # configuration.
  #
  # Whenever there is no content to be written into such a file, the key is
  # not used as filename - yet it can still be used as index for merging
  # configuration.
  #
  # The values inside the entries consist of the following optional entries:
  #   'source': a sources.list entry (some variable replacements apply)
  #   'keyid': providing a key to import via shortid or fingerprint
  #   'key': providing a raw PGP key
  #   'keyserver': specify an alternate keyserver to pull keys from that
  #                were specified by keyid

  # This allows merging between multiple input files than a list like:
  # cloud-config1
  # sources:
  #   s1: {'key': 'key1', 'source': 'source1'}
  # cloud-config2
  # sources:
  #   s2: {'key': 'key2'}
  #   s1: {'keyserver': 'foo'}
  # This would be merged to
  # sources:
  #   s1:
  #     keyserver: foo
  #     key: key1
  #     source: source1
  #   s2:
  #     key: key2
  #
  # The following examples number the subfeatures per sources entry to ease
  # identification in discussions.


  sources:
    curtin-dev-ppa.list:
      # 2.1 source
      #
      # Creates a file in /etc/apt/sources.list.d/ for the sources list entry
      # based on the key: "/etc/apt/sources.list.d/curtin-dev-ppa.list"
      source: "deb http://ppa.launchpad.net/curtin-dev/test-archive/ubuntu bionic main"

      # 2.2 keyid
      #
      # Importing a gpg key for a given key id. Used keyserver defaults to
      # keyserver.ubuntu.com
      keyid: F430BBA5 # GPG key ID published on a key server

    ignored1:
      # 2.3 PPA shortcut
      #
      # Setup correct apt sources.list line and Auto-Import the signing key
      # from LP
      #
      # See https://help.launchpad.net/Packaging/PPA for more information
      # this requires 'add-apt-repository'. This will create a file in
      # /etc/apt/sources.list.d automatically, therefore the key here is
      # ignored as filename in those cases.
      source: "ppa:curtin-dev/test-archive"    # Quote the string

    my-repo2.list:
      # 2.4 replacement variables
      #
      # sources can use $MIRROR, $PRIMARY, $SECURITY, $RELEASE and $KEY_FILE
      # replacement variables.
      # They will be replaced with the default or specified mirrors and the
      # running release.
      # The entry below would be possibly turned into:
      #   source: deb http://archive.ubuntu.com/ubuntu bionic multiverse
      source: deb [signed-by=$KEY_FILE] $MIRROR $RELEASE multiverse
      keyid: F430BBA5

    my-repo3.list:
      # this would have the same end effect as 'ppa:curtin-dev/test-archive'
      source: "deb http://ppa.launchpad.net/curtin-dev/test-archive/ubuntu bionic main"
      keyid: F430BBA5 # GPG key ID published on the key server
      filename: curtin-dev-ppa.list

    ignored2:
      # 2.5 key only
      #
      # this would only import the key without adding a ppa or other source spec
      # since this doesn't generate a source.list file the filename key is ignored
      keyid: F430BBA5 # GPG key ID published on a key server

    ignored3:
      # 2.6 key id alternatives
      #
      # Keyid's can also be specified via their long fingerprints
      keyid: B59D 5F15 97A5 04B7 E230  6DCA 0620 BBCF 0368 3F77

    ignored4:
      # 2.7 alternative keyservers
      #
      # One can also specify alternative keyservers to fetch keys from.
      keyid: B59D 5F15 97A5 04B7 E230  6DCA 0620 BBCF 0368 3F77
      keyserver: pgp.mit.edu

    ignored5:
      # 2.8 signed-by
      #
      # One can specify [signed-by=$KEY_FILE] in the source definition, which
      # will make the key be installed in the directory /etc/cloud-init.gpg.d/
      # and the $KEY_FILE replacement variable will be replaced with the path
      # to the specified key. If $KEY_FILE is used, but no key is specified,
      # apt update will (rightfully) fail due to an invalid value.
      source: deb [signed-by=$KEY_FILE] $MIRROR $RELEASE multiverse
      keyid: B59D 5F15 97A5 04B7 E230  6DCA 0620 BBCF 0368 3F77

    my-repo4.list:
      # 2.9 raw key
      #
      # The apt signing key can also be specified by providing a pgp public key
      # block. Providing the PGP key this way is the most robust method for
      # specifying a key, as it removes dependency on a remote key server.
      #
      # As with keyid's this can be specified with or without some actual source
      # content.
      key: | # The value needs to start with -----BEGIN PGP PUBLIC KEY BLOCK-----
        -----BEGIN PGP PUBLIC KEY BLOCK-----
        Version: SKS 1.0.10

        mI0ESpA3UQEEALdZKVIMq0j6qWAXAyxSlF63SvPVIgxHPb9Nk0DZUixn+akqytxG4zKCONz6
        qLjoBBfHnynyVLfT4ihg9an1PqxRnTO+JKQxl8NgKGz6Pon569GtAOdWNKw15XKinJTDLjnj
        9y96ljJqRcpV9t/WsIcdJPcKFR5voHTEoABE2aEXABEBAAG0GUxhdW5jaHBhZCBQUEEgZm9y
        IEFsZXN0aWOItgQTAQIAIAUCSpA3UQIbAwYLCQgHAwIEFQIIAwQWAgMBAh4BAheAAAoJEA7H
        5Qi+CcVxWZ8D/1MyYvfj3FJPZUm2Yo1zZsQ657vHI9+pPouqflWOayRR9jbiyUFIn0VdQBrP
        t0FwvnOFArUovUWoKAEdqR8hPy3M3APUZjl5K4cMZR/xaMQeQRZ5CHpS4DBKURKAHC0ltS5o
        uBJKQOZm5iltJp15cgyIkBkGe8Mx18VFyVglAZey
        =Y2oI
        -----END PGP PUBLIC KEY BLOCK-----

Disk setup

#cloud-config
# Cloud-init supports the creation of simple partition tables and file systems
# on devices.

# Default disk definitions for AWS
# --------------------------------
# (Not implemented yet, but provided for future documentation)

disk_setup:
  ephmeral0:
    table_type: 'mbr'
    layout: True
    overwrite: False

fs_setup:
  - label: None,
    filesystem: ext3
    device: ephemeral0
    partition: auto

# Default disk definitions for Microsoft Azure
# ------------------------------------------

device_aliases: {'ephemeral0': '/dev/sdb'}
disk_setup:
  ephemeral0:
    table_type: mbr
    layout: True
    overwrite: False

fs_setup:
  - label: ephemeral0
    filesystem: ext4
    device: ephemeral0.1
    replace_fs: ntfs


# Data disks definitions for Microsoft Azure
# ------------------------------------------

disk_setup:
  /dev/disk/azure/scsi1/lun0:
    table_type: gpt
    layout: True
    overwrite: True

fs_setup:
  - device: /dev/disk/azure/scsi1/lun0
    partition: 1
    filesystem: ext4


# Default disk definitions for SmartOS
# ------------------------------------

device_aliases: {'ephemeral0': '/dev/vdb'}
disk_setup:
  ephemeral0:
    table_type: mbr
    layout: False
    overwrite: False

fs_setup:
  - label: ephemeral0
    filesystem: ext4
    device: ephemeral0.0

# Caveat for SmartOS: if ephemeral disk is not defined, then the disk will
#    not be automatically added to the mounts.


# The default definition is used to make sure that the ephemeral storage is
# setup properly.

# "disk_setup": disk partitioning
# --------------------------------

# The disk_setup directive instructs Cloud-init to partition a disk. The format is:

disk_setup:
  ephmeral0:
    table_type: 'mbr'
    layout: true
  /dev/xvdh:
    table_type: 'mbr'
    layout:
      - 33
      - [33, 82]
      - 33
    overwrite: True

# The format is a list of dicts of dicts. The first value is the name of the
# device and the subsequent values define how to create and layout the
# partition.
# The general format is:
#   disk_setup:
#     <DEVICE>:
#       table_type: 'mbr'
#       layout: <LAYOUT|BOOL>
#       overwrite: <BOOL>
#
# Where:
#   <DEVICE>: The name of the device. 'ephemeralX' and 'swap' are special
#               values which are specific to the cloud. For these devices
#               Cloud-init will look up what the real devices is and then
#               use it.
#
#               For other devices, the kernel device name is used. At this
#               time only simply kernel devices are supported, meaning
#               that device mapper and other targets may not work.
#
#               Note: At this time, there is no handling or setup of
#               device mapper targets.
#
#   table_type=<TYPE>: Currently the following are supported:
#                   'mbr': default and setups a MS-DOS partition table
#                   'gpt': setups a GPT partition table
#
#               Note: At this time only 'mbr' and 'gpt' partition tables
#                   are allowed. It is anticipated in the future that
#                   we'll also have "RAID" to create a mdadm RAID.
#
#   layout={...}: The device layout. This is a list of values, with the
#               percentage of disk that partition will take.
#               Valid options are:
#                   [<SIZE>, [<SIZE>, <PART_TYPE]]
#
#               Where <SIZE> is the _percentage_ of the disk to use, while
#               <PART_TYPE> is the numerical value of the partition type.
#
#               The following setups two partitions, with the first
#               partition having a swap label, taking 1/3 of the disk space
#               and the remainder being used as the second partition.
#                 /dev/xvdh':
#                   table_type: 'mbr'
#                   layout:
#                     - [33,82]
#                     - 66
#                   overwrite: True
#
#               When layout is "true" it means single partition the entire
#               device.
#
#               When layout is "false" it means don't partition or ignore
#               existing partitioning.
#
#               If layout is set to "true" and overwrite is set to "false",
#               it will skip partitioning the device without a failure.
#
#   overwrite=<BOOL>: This describes whether to ride with saftey's on and
#               everything holstered.
#
#               'false' is the default, which means that:
#                   1. The device will be checked for a partition table
#                   2. The device will be checked for a file system
#                   3. If either a partition of file system is found, then
#                       the operation will be _skipped_.
#
#               'true' is cowboy mode. There are no checks and things are
#                   done blindly. USE with caution, you can do things you
#                   really, really don't want to do.
#
#
# fs_setup: Setup the file system
# -------------------------------
#
# fs_setup describes the how the file systems are supposed to look.

fs_setup:
  - label: ephemeral0
    filesystem: 'ext3'
    device: 'ephemeral0'
    partition: 'auto'
  - label: mylabl2
    filesystem: 'ext4'
    device: '/dev/xvda1'
  - cmd: mkfs -t %(filesystem)s -L %(label)s %(device)s
    label: mylabl3
    filesystem: 'btrfs'
    device: '/dev/xvdh'

# The general format is:
#   fs_setup:
#     - label: <LABEL>
#       filesystem: <FS_TYPE>
#       device: <DEVICE>
#       partition: <PART_VALUE>
#       overwrite: <OVERWRITE>
#       replace_fs: <FS_TYPE>
#
# Where:
#   <LABEL>: The file system label to be used. If set to None, no label is
#     used.
#
#   <FS_TYPE>: The file system type. It is assumed that the there
#     will be a "mkfs.<FS_TYPE>" that behaves likes "mkfs". On a standard
#     Ubuntu Cloud Image, this means that you have the option of ext{2,3,4},
#     and vfat by default.
#
#   <DEVICE>: The device name. Special names of 'ephemeralX' or 'swap'
#     are allowed and the actual device is acquired from the cloud datasource.
#     When using 'ephemeralX' (i.e. ephemeral0), make sure to leave the
#     label as 'ephemeralX' otherwise there may be issues with the mounting
#     of the ephemeral storage layer.
#
#     If you define the device as 'ephemeralX.Y' then Y will be interpetted
#     as a partition value. However, ephermalX.0 is the _same_ as ephemeralX.
#
#   <PART_VALUE>:
#     Partition definitions are overwritten if you use the '<DEVICE>.Y' notation.
#
#     The valid options are:
#     "auto|any": tell cloud-init not to care whether there is a partition
#       or not. Auto will use the first partition that does not contain a
#       file system already. In the absence of a partition table, it will
#       put it directly on the disk.
#
#       "auto": If a file system that matches the specification in terms of
#       label, type and device, then cloud-init will skip the creation of
#       the file system.
#
#       "any": If a file system that matches the file system type and device,
#       then cloud-init will skip the creation of the file system.
#
#       Devices are selected based on first-detected, starting with partitions
#       and then the raw disk. Consider the following:
#           NAME     FSTYPE LABEL
#           xvdb
#           |-xvdb1  ext4
#           |-xvdb2
#           |-xvdb3  btrfs  test
#           \-xvdb4  ext4   test
#
#         If you ask for 'auto', label of 'test, and file system of 'ext4'
#         then cloud-init will select the 2nd partition, even though there
#         is a partition match at the 4th partition.
#
#         If you ask for 'any' and a label of 'test', then cloud-init will
#         select the 1st partition.
#
#         If you ask for 'auto' and don't define label, then cloud-init will
#         select the 1st partition.
#
#         In general, if you have a specific partition configuration in mind,
#         you should define either the device or the partition number. 'auto'
#         and 'any' are specifically intended for formatting ephemeral storage or
#         for simple schemes.
#
#       "none": Put the file system directly on the device.
#
#       <NUM>: where NUM is the actual partition number.
#
#   <OVERWRITE>: Defines whether or not to overwrite any existing
#     filesystem.
#
#     "true": Indiscriminately destroy any pre-existing file system. Use at
#         your own peril.
#
#     "false": If an existing file system exists, skip the creation.
#
#   <REPLACE_FS>: This is a special directive, used for Microsoft Azure that
#     instructs cloud-init to replace a file system of <FS_TYPE>. NOTE:
#     unless you define a label, this requires the use of the 'any' partition
#     directive.
#
# Behavior Caveat: The default behavior is to _check_ if the file system exists.
#   If a file system matches the specification, then the operation is a no-op.

Configure data sources

#cloud-config

# Documentation on data sources configuration options
datasource:
  # Ec2 
  Ec2:
    # timeout: the timeout value for a request at metadata service
    timeout : 50
    # The length in seconds to wait before giving up on the metadata
    # service.  The actual total wait could be up to 
    #   len(resolvable_metadata_urls)*timeout
    max_wait : 120

    #metadata_url: a list of URLs to check for metadata services
    metadata_urls:
     - http://169.254.169.254:80
     - http://instance-data:8773

  MAAS:
    timeout : 50
    max_wait : 120

    # there are no default values for metadata_url or oauth credentials
    # If no credentials are present, non-authed attempts will be made.
    metadata_url: http://mass-host.localdomain/source
    consumer_key: Xh234sdkljf
    token_key: kjfhgb3n
    token_secret: 24uysdfx1w4

  NoCloud:
    # default seedfrom is None
    # if found, then it should contain a url with:
    #    <url>/user-data and <url>/meta-data
    # seedfrom: http://my.example.com/i-abcde
    seedfrom: None

    # fs_label: the label on filesystems to be searched for NoCloud source
    fs_label: cidata

    # these are optional, but allow you to basically provide a datasource
    # right here
    user-data: |
      # This is the user-data verbatim
    meta-data:
      instance-id: i-87018aed
      local-hostname: myhost.internal

  SmartOS:
    # For KVM guests:
    # Smart OS datasource works over a serial console interacting with
    # a server on the other end. By default, the second serial console is the
    # device. SmartOS also uses a serial timeout of 60 seconds.
    serial_device: /dev/ttyS1
    serial_timeout: 60

    # For LX-Brand Zones guests:
    # Smart OS datasource works over a socket interacting with
    # the host on the other end. By default, the socket file is in
    # the native .zoncontrol directory.
    metadata_sockfile: /native/.zonecontrol/metadata.sock

    # a list of keys that will not be base64 decoded even if base64_all
    no_base64_decode: ['root_authorized_keys', 'motd_sys_info',
                       'iptables_disable']
    # a plaintext, comma delimited list of keys whose values are b64 encoded
    base64_keys: []
    # a boolean indicating that all keys not in 'no_base64_decode' are encoded
    base64_all: False

Create partitions and filesystems

#cloud-config
# Cloud-init supports the creation of simple partition tables and file systems
# on devices.

# Default disk definitions for AWS
# --------------------------------
# (Not implemented yet, but provided for future documentation)

disk_setup:
  ephmeral0:
    table_type: 'mbr'
    layout: True
    overwrite: False

fs_setup:
  - label: None,
    filesystem: ext3
    device: ephemeral0
    partition: auto

# Default disk definitions for Microsoft Azure
# ------------------------------------------

device_aliases: {'ephemeral0': '/dev/sdb'}
disk_setup:
  ephemeral0:
    table_type: mbr
    layout: True
    overwrite: False

fs_setup:
  - label: ephemeral0
    filesystem: ext4
    device: ephemeral0.1
    replace_fs: ntfs


# Data disks definitions for Microsoft Azure
# ------------------------------------------

disk_setup:
  /dev/disk/azure/scsi1/lun0:
    table_type: gpt
    layout: True
    overwrite: True

fs_setup:
  - device: /dev/disk/azure/scsi1/lun0
    partition: 1
    filesystem: ext4


# Default disk definitions for SmartOS
# ------------------------------------

device_aliases: {'ephemeral0': '/dev/vdb'}
disk_setup:
  ephemeral0:
    table_type: mbr
    layout: False
    overwrite: False

fs_setup:
  - label: ephemeral0
    filesystem: ext4
    device: ephemeral0.0

# Caveat for SmartOS: if ephemeral disk is not defined, then the disk will
#    not be automatically added to the mounts.


# The default definition is used to make sure that the ephemeral storage is
# setup properly.

# "disk_setup": disk partitioning
# --------------------------------

# The disk_setup directive instructs Cloud-init to partition a disk. The format is:

disk_setup:
  ephmeral0:
    table_type: 'mbr'
    layout: true
  /dev/xvdh:
    table_type: 'mbr'
    layout:
      - 33
      - [33, 82]
      - 33
    overwrite: True

# The format is a list of dicts of dicts. The first value is the name of the
# device and the subsequent values define how to create and layout the
# partition.
# The general format is:
#   disk_setup:
#     <DEVICE>:
#       table_type: 'mbr'
#       layout: <LAYOUT|BOOL>
#       overwrite: <BOOL>
#
# Where:
#   <DEVICE>: The name of the device. 'ephemeralX' and 'swap' are special
#               values which are specific to the cloud. For these devices
#               Cloud-init will look up what the real devices is and then
#               use it.
#
#               For other devices, the kernel device name is used. At this
#               time only simply kernel devices are supported, meaning
#               that device mapper and other targets may not work.
#
#               Note: At this time, there is no handling or setup of
#               device mapper targets.
#
#   table_type=<TYPE>: Currently the following are supported:
#                   'mbr': default and setups a MS-DOS partition table
#                   'gpt': setups a GPT partition table
#
#               Note: At this time only 'mbr' and 'gpt' partition tables
#                   are allowed. It is anticipated in the future that
#                   we'll also have "RAID" to create a mdadm RAID.
#
#   layout={...}: The device layout. This is a list of values, with the
#               percentage of disk that partition will take.
#               Valid options are:
#                   [<SIZE>, [<SIZE>, <PART_TYPE]]
#
#               Where <SIZE> is the _percentage_ of the disk to use, while
#               <PART_TYPE> is the numerical value of the partition type.
#
#               The following setups two partitions, with the first
#               partition having a swap label, taking 1/3 of the disk space
#               and the remainder being used as the second partition.
#                 /dev/xvdh':
#                   table_type: 'mbr'
#                   layout:
#                     - [33,82]
#                     - 66
#                   overwrite: True
#
#               When layout is "true" it means single partition the entire
#               device.
#
#               When layout is "false" it means don't partition or ignore
#               existing partitioning.
#
#               If layout is set to "true" and overwrite is set to "false",
#               it will skip partitioning the device without a failure.
#
#   overwrite=<BOOL>: This describes whether to ride with saftey's on and
#               everything holstered.
#
#               'false' is the default, which means that:
#                   1. The device will be checked for a partition table
#                   2. The device will be checked for a file system
#                   3. If either a partition of file system is found, then
#                       the operation will be _skipped_.
#
#               'true' is cowboy mode. There are no checks and things are
#                   done blindly. USE with caution, you can do things you
#                   really, really don't want to do.
#
#
# fs_setup: Setup the file system
# -------------------------------
#
# fs_setup describes the how the file systems are supposed to look.

fs_setup:
  - label: ephemeral0
    filesystem: 'ext3'
    device: 'ephemeral0'
    partition: 'auto'
  - label: mylabl2
    filesystem: 'ext4'
    device: '/dev/xvda1'
  - cmd: mkfs -t %(filesystem)s -L %(label)s %(device)s
    label: mylabl3
    filesystem: 'btrfs'
    device: '/dev/xvdh'

# The general format is:
#   fs_setup:
#     - label: <LABEL>
#       filesystem: <FS_TYPE>
#       device: <DEVICE>
#       partition: <PART_VALUE>
#       overwrite: <OVERWRITE>
#       replace_fs: <FS_TYPE>
#
# Where:
#   <LABEL>: The file system label to be used. If set to None, no label is
#     used.
#
#   <FS_TYPE>: The file system type. It is assumed that the there
#     will be a "mkfs.<FS_TYPE>" that behaves likes "mkfs". On a standard
#     Ubuntu Cloud Image, this means that you have the option of ext{2,3,4},
#     and vfat by default.
#
#   <DEVICE>: The device name. Special names of 'ephemeralX' or 'swap'
#     are allowed and the actual device is acquired from the cloud datasource.
#     When using 'ephemeralX' (i.e. ephemeral0), make sure to leave the
#     label as 'ephemeralX' otherwise there may be issues with the mounting
#     of the ephemeral storage layer.
#
#     If you define the device as 'ephemeralX.Y' then Y will be interpetted
#     as a partition value. However, ephermalX.0 is the _same_ as ephemeralX.
#
#   <PART_VALUE>:
#     Partition definitions are overwritten if you use the '<DEVICE>.Y' notation.
#
#     The valid options are:
#     "auto|any": tell cloud-init not to care whether there is a partition
#       or not. Auto will use the first partition that does not contain a
#       file system already. In the absence of a partition table, it will
#       put it directly on the disk.
#
#       "auto": If a file system that matches the specification in terms of
#       label, type and device, then cloud-init will skip the creation of
#       the file system.
#
#       "any": If a file system that matches the file system type and device,
#       then cloud-init will skip the creation of the file system.
#
#       Devices are selected based on first-detected, starting with partitions
#       and then the raw disk. Consider the following:
#           NAME     FSTYPE LABEL
#           xvdb
#           |-xvdb1  ext4
#           |-xvdb2
#           |-xvdb3  btrfs  test
#           \-xvdb4  ext4   test
#
#         If you ask for 'auto', label of 'test, and file system of 'ext4'
#         then cloud-init will select the 2nd partition, even though there
#         is a partition match at the 4th partition.
#
#         If you ask for 'any' and a label of 'test', then cloud-init will
#         select the 1st partition.
#
#         If you ask for 'auto' and don't define label, then cloud-init will
#         select the 1st partition.
#
#         In general, if you have a specific partition configuration in mind,
#         you should define either the device or the partition number. 'auto'
#         and 'any' are specifically intended for formatting ephemeral storage or
#         for simple schemes.
#
#       "none": Put the file system directly on the device.
#
#       <NUM>: where NUM is the actual partition number.
#
#   <OVERWRITE>: Defines whether or not to overwrite any existing
#     filesystem.
#
#     "true": Indiscriminately destroy any pre-existing file system. Use at
#         your own peril.
#
#     "false": If an existing file system exists, skip the creation.
#
#   <REPLACE_FS>: This is a special directive, used for Microsoft Azure that
#     instructs cloud-init to replace a file system of <FS_TYPE>. NOTE:
#     unless you define a label, this requires the use of the 'any' partition
#     directive.
#
# Behavior Caveat: The default behavior is to _check_ if the file system exists.
#   If a file system matches the specification, then the operation is a no-op.

Grow partitions

#cloud-config
#
# growpart entry is a dict, if it is not present at all
# in config, then the default is used ({'mode': 'auto', 'devices': ['/']})
#
#  mode:
#    values:
#     * auto: use any option possible (any available)
#             if none are available, do not warn, but debug.
#     * growpart: use growpart to grow partitions
#             if growpart is not available, this is an error.
#     * off, false
#
# devices:
#   a list of things to resize.
#   items can be filesystem paths or devices (in /dev)
#   examples:
#     devices: [/, /dev/vdb1]
#
# ignore_growroot_disabled:
#   a boolean, default is false.
#   if the file /etc/growroot-disabled exists, then cloud-init will not grow
#   the root partition.  This is to allow a single file to disable both
#   cloud-initramfs-growroot and cloud-init's growroot support.
#
#   true indicates that /etc/growroot-disabled should be ignored
#
growpart:
  mode: auto
  devices: ['/']
  ignore_growroot_disabled: false

Events and Updates

Events

Cloud-init will fetch and apply cloud and user data configuration upon several event types. The two most common events for cloud-init are when an instance first boots and any subsequent boot thereafter (reboot). In addition to boot events, cloud-init users and vendors are interested in when devices are added. cloud-init currently supports the following event types:

• BOOT_NEW_INSTANCE: New instance first boot

• BOOT: Any system boot other than ‘BOOT_NEW_INSTANCE’

• BOOT_LEGACY: Similar to ‘BOOT’, but applies networking config twice each boot: once during Local stage, then again in Network stage. As this behavior was previously the default behavior, this option exists to prevent regressing such behavior.

• HOTPLUG: Dynamic add of a system device

Future work will likely include infrastructure and support for the following events:

• METADATA_CHANGE: An instance’s metadata has change

• USER_REQUEST: Directed request to update

Datasource Event Support

All Datasources by default support the BOOT_NEW_INSTANCE event. Each Datasource will declare a set of these events that it is capable of handling. Datasources may not support all event types. In some cases a system may be configured to allow a particular event but may be running on a platform whose datasource cannot support the event.

Configuring Event Updates

Update configuration may be specified via user data, which can be used to enable or disable handling of specific events. This configuration will be honored as long as the events are supported by the datasource. However, configuration will always be applied at first boot, regardless of the user data specified.

Updates

Update policy configuration defines which events are allowed to be handled. This is separate from whether a particular platform or datasource has the capability for such events.

scope: <name of the scope for event policy>

The scope value is a string which defines under which domain does the event occur. Currently the only one known scope is network, though more scopes may be added in the future. Scopes are defined by convention but arbitrary values can be used.

when: <list of events to handle for a particular scope>

Each scope requires a when element to specify which events are to allowed to be handled.

Hotplug

When the hotplug event is supported by the data source and configured in user data, cloud-init will respond to the addition or removal of network interfaces to the system. In addition to fetching and updating the system metadata, cloud-init will also bring up/down the newly added interface.

Warning

Due to its use of systemd sockets, hotplug functionality is currently incompatible with SELinux. This issue is being tracked on Launchpad. Additionally, hotplug support is considered experimental for non-Debian based systems.

Examples

apply network config every boot

On every boot, apply network configuration found in the datasource.

# apply network config on every boot
updates:
  network:
    when: ['boot']

Modules

Table of Contents

• Modules

  • APK Configure

  • Apt Configure

  • Apt Pipelining

  • Bootcmd

  • Byobu

  • CA Certificates

  • Chef

  • Debug

  • Disable EC2 Metadata

  • Disk Setup

  • Emit Upstart

  • Fan

  • Final Message

  • Foo

  • Growpart

  • Grub Dpkg

  • Install Hotplug

  • Keyboard

  • Keys to Console

  • Landscape

  • Locale

  • LXD

  • Mcollective

  • Migrator

  • Mounts

  • NTP

  • Package Update Upgrade Install

  • Phone Home

  • Power State Change

  • Puppet

  • Resizefs

  • Resolv Conf

  • Red Hat Subscription

  • RightScale Userdata

  • Rsyslog

  • Runcmd

  • Salt Minion

  • Scripts Per Boot

  • Scripts Per Instance

  • Scripts Per Once

  • Scripts User

  • Scripts Vendor

  • Seed Random

  • Set Hostname

  • Set Passwords

  • Snap

  • Spacewalk

  • SSH

    • Authorized Keys

    • Host Keys

  • SSH Authkey Fingerprints

  • SSH Import Id

  • Timezone

  • Ubuntu Advantage

  • Ubuntu Drivers

  • Update Etc Hosts

  • Update Hostname

  • Users and Groups

  • Write Files

  • Yum Add Repo

APK Configure

Summary: Configure apk repositories file

This module handles configuration of the /etc/apk/repositories file.

Note

To ensure that apk configuration is valid yaml, any strings containing special characters, especially : should be quoted.

Internal name: cc_apk_configure

Module frequency: once-per-instance

Supported distros: alpine

Config schema:

apk_repos: (object)

preserve_repositories: (boolean) By default, cloud-init will generate a new repositories file /etc/apk/repositories based on any valid configuration settings specified within a apk_repos section of cloud config. To disable this behavior and preserve the repositories file from the pristine image, set preserve_repositories to true.

The preserve_repositories option overrides all other config keys that would alter /etc/apk/repositories.

alpine_repo: (object/null)

base_url: (string) The base URL of an Alpine repository, or mirror, to download official packages from. If not specified then it defaults to https://alpine.global.ssl.fastly.net/alpine

community_enabled: (boolean) Whether to add the Community repo to the repositories file. By default the Community repo is not included.

testing_enabled: (boolean) Whether to add the Testing repo to the repositories file. By default the Testing repo is not included. It is only recommended to use the Testing repo on a machine running the Edge version of Alpine as packages installed from Testing may have dependencies that conflict with those in non-Edge Main or Community repos.

version: (string) The Alpine version to use (e.g. v3.12 or edge)

local_repo_base_url: (string) The base URL of an Alpine repository containing unofficial packages

Examples:

# Keep the existing /etc/apk/repositories file unaltered.
apk_repos:
    preserve_repositories: true

# --- Example2 ---
# Create repositories file for Alpine v3.12 main and community
# using default mirror site.
apk_repos:
    alpine_repo:
        community_enabled: true
        version: 'v3.12'

# --- Example3 ---
# Create repositories file for Alpine Edge main, community, and
# testing using a specified mirror site and also a local repo.
apk_repos:
    alpine_repo:
        base_url: 'https://some-alpine-mirror/alpine'
        community_enabled: true
        testing_enabled: true
        version: 'edge'
    local_repo_base_url: 'https://my-local-server/local-alpine'

Apt Configure

Summary: Configure apt for the user

This module handles both configuration of apt options and adding source lists. There are configuration options such as apt_get_wrapper and apt_get_command that control how cloud-init invokes apt-get. These configuration options are handled on a per-distro basis, so consult documentation for cloud-init’s distro support for instructions on using these config options.

Note

To ensure that apt configuration is valid yaml, any strings containing special characters, especially : should be quoted.

Note

For more information about apt configuration, see the Additional apt configuration example.

Internal name: cc_apt_configure

Module frequency: once-per-instance

Supported distros: ubuntu, debian

Config schema:

apt: (object)

preserve_sources_list: (boolean) By default, cloud-init will generate a new sources list in /etc/apt/sources.list.d based on any changes specified in cloud config. To disable this behavior and preserve the sources list from the pristine image, set preserve_sources_list to true.

The preserve_sources_list option overrides all other config keys that would alter sources.list or sources.list.d, except for additional sources to be added to sources.list.d.

disable_suites: (array of string) Entries in the sources list can be disabled using disable_suites, which takes a list of suites to be disabled. If the string $RELEASE is present in a suite in the disable_suites list, it will be replaced with the release name. If a suite specified in disable_suites is not present in sources.list it will be ignored. For convenience, several aliases are provided for`` disable_suites``:

• updates => $RELEASE-updates

• backports => $RELEASE-backports

• security => $RELEASE-security

• proposed => $RELEASE-proposed

• release => $RELEASE.

When a suite is disabled using disable_suites, its entry in sources.list is not deleted; it is just commented out.

primary: (array of object) The primary and security archive mirrors can be specified using the primary and security keys, respectively. Both the primary and security keys take a list of configs, allowing mirrors to be specified on a per-architecture basis. Each config is a dictionary which must have an entry for arches, specifying which architectures that config entry is for. The keyword default applies to any architecture not explicitly listed. The mirror url can be specified with the uri key, or a list of mirrors to check can be provided in order, with the first mirror that can be resolved being selected. This allows the same configuration to be used in different environment, with different hosts used for a local APT mirror. If no mirror is provided by uri or search, search_dns may be used to search for dns names in the format <distro>-mirror in each of the following:

• fqdn of this host per cloud metadata,

• localdomain,

• domains listed in /etc/resolv.conf.

If there is a dns entry for <distro>-mirror, then it is assumed that there is a distro mirror at http://<distro>-mirror.<domain>/<distro>. If the primary key is defined, but not the security key, then then configuration for primary is also used for security. If search_dns is used for the security key, the search pattern will be <distro>-security-mirror.

Each mirror may also specify a key to import via any of the following optional keys:

• keyid: a key to import via shortid or fingerprint.

• key: a raw PGP key.

• keyserver: alternate keyserver to pull keyid key from.

If no mirrors are specified, or all lookups fail, then default mirrors defined in the datasource are used. If none are present in the datasource either the following defaults are used:

• primary => http://archive.ubuntu.com/ubuntu.

• security => http://security.ubuntu.com/ubuntu

Each item in primary list supports the following keys:

arches: (array of string)

uri: (string)

search: (array of string)

search_dns: (boolean)

keyid: (string)

key: (string)

keyserver: (string)

security: (array of object) Please refer to the primary config documentation

Each item in security list supports the following keys:

arches: (array of string)

uri: (string)

search: (array of string)

search_dns: (boolean)

keyid: (string)

key: (string)

keyserver: (string)

add_apt_repo_match: (string) All source entries in apt-sources that match regex in add_apt_repo_match will be added to the system using add-apt-repository. If add_apt_repo_match is not specified, it defaults to ^[\w-]+:\w

debconf_selections: (object) Debconf additional configurations can be specified as a dictionary under the debconf_selections config key, with each key in the dict representing a different set of configurations. The value of each key must be a string containing all the debconf configurations that must be applied. We will bundle all of the values and pass them to debconf-set-selections. Therefore, each value line must be a valid entry for debconf-set-selections, meaning that they must possess for distinct fields:

pkgname question type answer

Where:

• pkgname is the name of the package.

• question the name of the questions.

• type is the type of question.

• answer is the value used to answer the question.

For example: ippackage ippackage/ip string 127.0.01

^.+$: (string)

sources_list: (string) Specifies a custom template for rendering sources.list . If no sources_list template is given, cloud-init will use sane default. Within this template, the following strings will be replaced with the appropriate values:

• $MIRROR

• $RELEASE

• $PRIMARY

• $SECURITY

• $KEY_FILE

conf: (string) Specify configuration for apt, such as proxy configuration. This configuration is specified as a string. For multiline APT configuration, make sure to follow yaml syntax.

https_proxy: (string) More convenient way to specify https APT proxy. https proxy url is specified in the format https://[[user][:pass]@]host[:port]/.

http_proxy: (string) More convenient way to specify http APT proxy. http proxy url is specified in the format http://[[user][:pass]@]host[:port]/.

proxy: (string) Alias for defining a http APT proxy.

ftp_proxy: (string) More convenient way to specify ftp APT proxy. ftp proxy url is specified in the format ftp://[[user][:pass]@]host[:port]/.

sources: (object) Source list entries can be specified as a dictionary under the sources config key, with each key in the dict representing a different source file. The key of each source entry will be used as an id that can be referenced in other config entries, as well as the filename for the source’s configuration under /etc/apt/sources.list.d. If the name does not end with .list, it will be appended. If there is no configuration for a key in sources, no file will be written, but the key may still be referred to as an id in other sources entries.

Each entry under sources is a dictionary which may contain any of the following optional keys:

• source: a sources.list entry (some variable replacements apply).

• keyid: a key to import via shortid or fingerprint.

• key: a raw PGP key.

• keyserver: alternate keyserver to pull keyid key from.

• filename: specify the name of the list file

The source key supports variable replacements for the following strings:

• $MIRROR

• $PRIMARY

• $SECURITY

• $RELEASE

• $KEY_FILE

^.+$: (object)

source: (string)

keyid: (string)

key: (string)

keyserver: (string)

filename: (string)

Examples:

apt:
  preserve_sources_list: false
  disable_suites:
    - $RELEASE-updates
    - backports
    - $RELEASE
    - mysuite
  primary:
    - arches:
        - amd64
        - i386
        - default
      uri: 'http://us.archive.ubuntu.com/ubuntu'
      search:
        - 'http://cool.but-sometimes-unreachable.com/ubuntu'
        - 'http://us.archive.ubuntu.com/ubuntu'
      search_dns: false
    - arches:
        - s390x
        - arm64
      uri: 'http://archive-to-use-for-arm64.example.com/ubuntu'

  security:
    - arches:
        - default
      search_dns: true
  sources_list: |
      deb $MIRROR $RELEASE main restricted
      deb-src $MIRROR $RELEASE main restricted
      deb $PRIMARY $RELEASE universe restricted
      deb $SECURITY $RELEASE-security multiverse
  debconf_selections:
      set1: the-package the-package/some-flag boolean true
  conf: |
      APT {
          Get {
              Assume-Yes 'true';
              Fix-Broken 'true';
          }
      }
  proxy: 'http://[[user][:pass]@]host[:port]/'
  http_proxy: 'http://[[user][:pass]@]host[:port]/'
  ftp_proxy: 'ftp://[[user][:pass]@]host[:port]/'
  https_proxy: 'https://[[user][:pass]@]host[:port]/'
  sources:
      source1:
          keyid: 'keyid'
          keyserver: 'keyserverurl'
          source: 'deb [signed-by=$KEY_FILE] http://<url>/ bionic main'
      source2:
          source: 'ppa:<ppa-name>'
      source3:
          source: 'deb $MIRROR $RELEASE multiverse'
          key: |
              ------BEGIN PGP PUBLIC KEY BLOCK-------
              <key data>
              ------END PGP PUBLIC KEY BLOCK-------

Apt Pipelining

Summary: Configure apt pipelining

This module configures apt’s Acquite::http::Pipeline-Depth option, which controls how apt handles HTTP pipelining. It may be useful for pipelining to be disabled, because some web servers, such as S3 do not pipeline properly (LP: #948461).

Value configuration options for this module are:

• false (Default): disable pipelining altogether

• none, unchanged, or os: use distro default

• <number>: Manually specify pipeline depth. This is not recommended.

Internal name: cc_apt_pipelining

Module frequency: once-per-instance

Supported distros: ubuntu, debian

Config schema:

apt_pipelining: (integer/boolean/string)

Examples:

apt_pipelining: false
# --- Example2 ---
apt_pipelining: none
# --- Example3 ---
apt_pipelining: unchanged
# --- Example4 ---
apt_pipelining: os
# --- Example5 ---
apt_pipelining: 3

Bootcmd

Summary: Run arbitrary commands early in the boot process

This module runs arbitrary commands very early in the boot process, only slightly after a boothook would run. This is very similar to a boothook, but more user friendly. The environment variable INSTANCE_ID will be set to the current instance id for all run commands. Commands can be specified either as lists or strings. For invocation details, see runcmd.

Note

bootcmd should only be used for things that could not be done later in the boot process.

Note

when writing files, do not use /tmp dir as it races with systemd-tmpfiles-clean LP: #1707222. Use /run/somedir instead.

Internal name: cc_bootcmd

Module frequency: always

Supported distros: all

Config schema:

bootcmd: (array of (array of string)/(string))

Examples:

bootcmd:
    - echo 192.168.1.130 us.archive.ubuntu.com > /etc/hosts
    - [ cloud-init-per, once, mymkfs, mkfs, /dev/vdb ]

Byobu

Summary: Enable/disable byobu system wide and for default user

This module controls whether byobu is enabled or disabled system wide and for the default system user. If byobu is to be enabled, this module will ensure it is installed. Likewise, if it is to be disabled, it will be removed if installed.

Valid configuration options for this module are:

• enable-system: enable byobu system wide

• enable-user: enable byobu for the default user

• disable-system: disable byobu system wide

• disable-user: disable byobu for the default user

• enable: enable byobu both system wide and for default user

• disable: disable byobu for all users

• user: alias for enable-user

• system: alias for enable-system

Internal name: cc_byobu

Module frequency: once-per-instance

Supported distros: ubuntu, debian

Config schema:

byobu_by_default: (string)

Examples:

byobu_by_default: enable-user
# --- Example2 ---
byobu_by_default: disable-system

CA Certificates

Summary: Add ca certificates

This module adds CA certificates to /etc/ca-certificates.conf and updates the ssl cert cache using update-ca-certificates. The default certificates can be removed from the system with the configuration option remove_defaults.

Note

certificates must be specified using valid yaml. in order to specify a multiline certificate, the yaml multiline list syntax must be used

Note

For Alpine Linux the “remove_defaults” functionality works if the ca-certificates package is installed but not if the ca-certificates-bundle package is installed.

Internal name: cc_ca_certs

Module frequency: once-per-instance

Supported distros: alpine, debian, ubuntu, rhel

Config schema:

ca_certs: (object)

remove-defaults: (boolean) DEPRECATED. Use remove_defaults.

remove_defaults: (boolean) Remove default CA certificates if true. Default: false

trusted: (array of string) List of trusted CA certificates to add.

ca-certs: (object)

remove-defaults: (boolean) DEPRECATED. Use remove_defaults.

remove_defaults: (boolean) Remove default CA certificates if true. Default: false

trusted: (array of string) List of trusted CA certificates to add.

Examples:

ca_certs:
  remove_defaults: true
  trusted:
    - single_line_cert
    - |
      -----BEGIN CERTIFICATE-----
      YOUR-ORGS-TRUSTED-CA-CERT-HERE
      -----END CERTIFICATE-----

Chef

Summary: module that configures, starts and installs chef

This module enables chef to be installed (from packages, gems, or from omnibus). Before this occurs, chef configuration is written to disk (validation.pem, client.pem, firstboot.json, client.rb), and required directories are created (/etc/chef and /var/log/chef and so-on). If configured, chef will be installed and started in either daemon or non-daemon mode. If run in non-daemon mode, post run actions are executed to do finishing activities such as removing validation.pem.

Internal name: cc_chef

Module frequency: always

Supported distros: all

Config schema:

chef: (object)

directories: (array of string) Create the necessary directories for chef to run. By default, it creates the following directories:

• /etc/chef

• /var/log/chef

• /var/lib/chef

• /var/cache/chef

• /var/backups/chef

• /var/run/chef

validation_cert: (string) Optional string to be written to file validation_key. Special value system means set use existing file.

validation_key: (string) Optional path for validation_cert. default to /etc/chef/validation.pem

firstboot_path: (string) Path to write run_list and initial_attributes keys that should also be present in this configuration, defaults to /etc/chef/firstboot.json

exec: (boolean) Set true if we should run or not run chef (defaults to false, unless a gem installed is requested where this will then default to true).

client_key: (string) Optional path for client_cert. Default to /etc/chef/client.pem.

encrypted_data_bag_secret: (string) Specifies the location of the secret key used by chef to encrypt data items. By default, this path is set to null, meaning that chef will have to look at the path /etc/chef/encrypted_data_bag_secret for it.

environment: (string) Specifies which environment chef will use. By default, it will use the _default configuration.

file_backup_path: (string) Specifies the location in which backup files are stored. By default, it uses the /var/backups/chef location.

file_cache_path: (string) Specifies the location in which chef cache files will be saved. By default, it uses the /var/cache/chef location.

json_attribs: (string) Specifies the location in which some chef json data is stored. By default, it uses the /etc/chef/firstboot.json location.

log_level: (string) Defines the level of logging to be stored in the log file. By default this value is set to :info.

log_location: (string) Specifies the location of the chef lof file. By default, the location is specified at /var/log/chef/client.log.

node_name: (string) The name of the node to run. By default, we will use th instance id as the node name.

omnibus_url: (string) Omnibus URL if chef should be installed through Omnibus. By default, it uses the https://www.chef.io/chef/install.sh.

omnibus_url_retries: (integer) The number of retries that will be attempted to reach the Omnibus URL. Default is 5.

omnibus_version: (string) Optional version string to require for omnibus install.

pid_file: (string) The location in which a process identification number (pid) is saved. By default, it saves in the /var/run/chef/client.pid location.

server_url: (string) The URL for the chef server

show_time: (boolean) Show time in chef logs

ssl_verify_mode: (string) Set the verify mode for HTTPS requests. We can have two possible values for this parameter:

• :verify_none: No validation of SSL certificates.

• :verify_peer: Validate all SSL certificates.

By default, the parameter is set as :verify_none.

validation_name: (string) The name of the chef-validator key that Chef Infra Client uses to access the Chef Infra Server during the initial Chef Infra Client run.

force_install: (boolean) If set to true, forces chef installation, even if it is already installed.

initial_attributes: (object of string) Specify a list of initial attributes used by the cookbooks.

install_type: (string) The type of installation for chef. It can be one of the following values:

• packages

• gems

• omnibus

run_list: (array of string) A run list for a first boot json.

chef_license: (string) string that indicates if user accepts or not license related to some of chef products

Examples:

chef:
  directories:
    - /etc/chef
    - /var/log/chef
  validation_cert: system
  install_type: omnibus
  initial_attributes:
    apache:
      prefork:
        maxclients: 100
      keepalive: off
  run_list:
    - recipe[apache2]
    - role[db]
  encrypted_data_bag_secret: /etc/chef/encrypted_data_bag_secret
  environment: _default
  log_level: :auto
  omnibus_url_retries: 2
  server_url: https://chef.yourorg.com:4000
  ssl_verify_mode: :verify_peer
  validation_name: yourorg-validator

Debug

Summary: Helper to debug cloud-init internal datastructures

This module will enable for outputting various internal information that cloud-init sources provide to either a file or to the output console/log location that this cloud-init has been configured with when running.

Note

Log configurations are not output.

Internal name: cc_debug

Module frequency: once-per-instance

Supported distros: all

Config schema:

debug: (object)

verbose: (boolean) Should always be true for this module

output: (string) Location to write output. Defaults to console + log

Examples:

debug:
  verbose: true
  output: /tmp/my_debug.log

Disable EC2 Metadata

Summary: Disable AWS EC2 Metadata

This module can disable the ec2 datasource by rejecting the route to 169.254.169.254, the usual route to the datasource. This module is disabled by default.

Internal name: cc_disable_ec2_metadata

Module frequency: always

Supported distros: all

Config schema:

disable_ec2_metadata: (boolean) Set true to disable IPv4 routes to EC2 metadata. Default: false.

Examples:

disable_ec2_metadata: true

Disk Setup

Summary: Configure partitions and filesystems

This module is able to configure simple partition tables and filesystems.

Note

for more detail about configuration options for disk setup, see the disk setup example

For convenience, aliases can be specified for disks using the device_aliases config key, which takes a dictionary of alias: path mappings. There are automatic aliases for swap and ephemeral<X>, where swap will always refer to the active swap partition and ephemeral<X> will refer to the block device of the ephemeral image.

Disk partitioning is done using the disk_setup directive. This config directive accepts a dictionary where each key is either a path to a block device or an alias specified in device_aliases, and each value is the configuration options for the device. File system configuration is done using the fs_setup directive. This config directive accepts a list of filesystem configs.

Internal name: cc_disk_setup

Module frequency: once-per-instance

Supported distros: all

Config schema:

device_aliases: (object)

<alias_name>: (string) Path to disk to be aliased by this name.

disk_setup: (object)

<alias name/path>: (object)

table_type: (string) Specifies the partition table type, either mbr or gpt. Default: mbr.

layout: (string/boolean/array) If set to true, a single partition using all the space on the device will be created. If set to false, no partitions will be created. If set to remove, any existing partition table will be purged. Partitions can be specified by providing a list to layout, where each entry in the list is either a size or a list containing a size and the numerical value for a partition type. The size for partitions is specified in percentage of disk space, not in bytes (e.g. a size of 33 would take up 1/3 of the disk space). Default: false.

overwrite: (boolean) Controls whether this module tries to be safe about writing partition tables or not. If overwrite: false is set, the device will be checked for a partition table and for a file system and if either is found, the operation will be skipped. If overwrite: true is set, no checks will be performed. Using overwrite: true is dangerous and can lead to data loss, so double check that the correct device has been specified if using this option. Default: false

fs_setup: (array of object)

Each item in fs_setup list supports the following keys:

label: (string) Label for the filesystem.

filesystem: (string) Filesystem type to create. E.g., ext4 or btrfs

device: (string) Specified either as a path or as an alias in the format <alias name>.<y> where <y> denotes the partition number on the device. If specifying device using the <device name>.<partition number> format, the value of partition will be overwritten.

partition: (string/integer) The partition can be specified by setting partition to the desired partition number. The partition option may also be set to auto, in which this module will search for the existence of a filesystem matching the label, type and device of the fs_setup entry and will skip creating the filesystem if one is found. The partition option may also be set to any, in which case any file system that matches type and device will cause this module to skip filesystem creation for the fs_setup entry, regardless of label matching or not. To write a filesystem directly to a device, use partition: none. partition: none will always write the filesystem, even when the label and filesystem are matched, and overwrite is false.

overwrite: (boolean) If true, overwrite any existing filesystem. Using overwrite: true for filesystems is dangerous and can lead to data loss, so double check the entry in fs_setup. Default: false

replace_fs: (string) Ignored unless partition is auto or any. Default false.

extra_opts: (array/string of string) Optional options to pass to the filesystem creation command. Ignored if you using cmd directly.

cmd: (array/string of string) Optional command to run to create the filesystem. Can include string substitutions of the other fs_setup config keys. This is only necessary if you need to override the default command.

Examples:

device_aliases:
  my_alias: /dev/sdb
disk_setup:
  my_alias:
    table_type: gpt
    layout: [50, 50]
    overwrite: true
fs_setup:
- label: fs1
  filesystem: ext4
  device: my_alias.1
  cmd: mkfs -t %(filesystem)s -L %(label)s %(device)s
- label: fs2
  device: my_alias.2
  filesystem: ext4
mounts:
- ["my_alias.1", "/mnt1"]
- ["my_alias.2", "/mnt2"]

Emit Upstart

Summary: emit upstart configuration

Emit upstart configuration for cloud-init modules on upstart based systems. No user configuration should be required.

Internal name: cc_emit_upstart

Module frequency: always

Supported distros: ubuntu, debian

Fan

Summary: configure ubuntu fan networking

This module installs, configures and starts the ubuntu fan network system. For more information about Ubuntu Fan, see: https://wiki.ubuntu.com/FanNetworking.

If cloud-init sees a fan entry in cloud-config it will:

• write config_path with the contents of the config key

• install the package ubuntu-fan if it is not installed

• ensure the service is started (or restarted if was previously running)

Internal name: cc_fan

Module frequency: per instance

Supported distros: ubuntu

Config keys:

fan:
    config: |
        # fan 240
        10.0.0.0/8 eth0/16 dhcp
        10.0.0.0/8 eth1/16 dhcp off
        # fan 241
        241.0.0.0/8 eth0/16 dhcp
    config_path: /etc/network/fan

Final Message

Summary: output final message when cloud-init has finished

This module configures the final message that cloud-init writes. The message is specified as a jinja template with the following variables set:

• version: cloud-init version

• timestamp: time at cloud-init finish

• datasource: cloud-init data source

• uptime: system uptime

Internal name: cc_final_message

Module frequency: always

Supported distros: all

Config keys:

final_message: <message>

Foo

Summary: example module

Example to show module structure. Does not do anything.

Internal name: cc_foo

Module frequency: per instance

Supported distros: all

Growpart

Summary: grow partitions

Growpart resizes partitions to fill the available disk space. This is useful for cloud instances with a larger amount of disk space available than the pristine image uses, as it allows the instance to automatically make use of the extra space.

The devices on which to run growpart are specified as a list under the devices key. Each entry in the devices list can be either the path to the device’s mountpoint in the filesystem or a path to the block device in /dev.

The utility to use for resizing can be selected using the mode config key. If the mode key is set to auto, then any available utility (either growpart or BSD gpart) will be used. If neither utility is available, no error will be raised. If mode is set to growpart, then the growpart utility will be used. If this utility is not available on the system, this will result in an error. If mode is set to off or false, then cc_growpart will take no action.

There is some functionality overlap between this module and the growroot functionality of cloud-initramfs-tools. However, there are some situations where one tool is able to function and the other is not. The default configuration for both should work for most cloud instances. To explicitly prevent cloud-initramfs-tools from running growroot, the file /etc/growroot-disabled can be created. By default, both growroot and cc_growpart will check for the existence of this file and will not run if it is present. However, this file can be ignored for cc_growpart by setting ignore_growroot_disabled to true. For more information on cloud-initramfs-tools see: https://launchpad.net/cloud-initramfs-tools

Growpart is enabled by default on the root partition. The default config for growpart is:

growpart:
    mode: auto
    devices: ["/"]
    ignore_growroot_disabled: false

Internal name: cc_growpart

Module frequency: always

Supported distros: all

Config keys:

growpart:
    mode: <auto/growpart/off/false>
    devices:
        - "/"
        - "/dev/vdb1"
    ignore_growroot_disabled: <true/false>

Grub Dpkg

Summary: configure grub debconf installation device

Configure which device is used as the target for grub installation. This module should work correctly by default without any user configuration. It can be enabled/disabled using the enabled config key in the grub_dpkg config dict. The global config key grub-dpkg is an alias for grub_dpkg. If no installation device is specified this module will execute grub-probe to determine which disk the /boot directory is associated with.

The value which is placed into the debconf database is in the format which the grub postinstall script expects. Normally, this is a /dev/disk/by-id/ value, but we do fallback to the plain disk name if a by-id name is not present.

If this module is executed inside a container, then the debconf database is seeded with empty values, and install_devices_empty is set to true.

Internal name: cc_grub_dpkg

Module frequency: per instance

Supported distros: ubuntu, debian

Config keys:

grub_dpkg:
    enabled: <true/false>
    grub-pc/install_devices: <devices>
    grub-pc/install_devices_empty: <devices>
grub-dpkg: (alias for grub_dpkg)

Install Hotplug

Summary: Install hotplug if supported and enabled

This module will install the udev rules to enable hotplug if supported by the datasource and enabled in the userdata. The udev rules will be installed as /etc/udev/rules.d/10-cloud-init-hook-hotplug.rules.

When hotplug is enabled, newly added network devices will be added to the system by cloud-init. After udev detects the event, cloud-init will referesh the instance metadata from the datasource, detect the device in the updated metadata, then apply the updated network configuration.

Currently supported datasources: Openstack, EC2

Internal name: cc_install_hotplug

Module frequency: once-per-instance

Supported distros: all

Config schema:

updates: (object)

network: (object)

when: (array of string)

Examples:

# Enable hotplug of network devices
updates:
  network:
    when: ["hotplug"]

# --- Example2 ---
# Enable network hotplug alongside boot event
updates:
  network:
    when: ["boot", "hotplug"]

Keyboard

Summary: Set keyboard layout

Handle keyboard configuration.

Internal name: cc_keyboard

Module frequency: once-per-instance

Supported distros: arch, debian, ubuntu, almalinux, amazon, centos, cloudlinux, eurolinux, fedora, miraclelinux, openEuler, photon, rhel, rocky, virtuozzo, opensuse, sles

Config schema:

keyboard: (object)

layout: (string) Required. Keyboard layout. Corresponds to XKBLAYOUT.

model: (string) Optional. Keyboard model. Corresponds to XKBMODEL.

variant: (string) Optional. Keyboard variant. Corresponds to XKBVARIANT.

options: (string) Optional. Keyboard options. Corresponds to XKBOPTIONS.

Examples:

# Set keyboard layout to "us"
keyboard:
  layout: us

# --- Example2 ---
# Set specific keyboard layout, model, variant, options
keyboard:
  layout: de
  model: pc105
  variant: nodeadkeys
  options: compose:rwin

Keys to Console

Summary: control which SSH host keys may be written to console

For security reasons it may be desirable not to write SSH host keys and their fingerprints to the console. To avoid either being written to the console the emit_keys_to_console config key under the main ssh config key can be used. To avoid the fingerprint of types of SSH host keys being written to console the ssh_fp_console_blacklist config key can be used. By default all types of keys will have their fingerprints written to console. To avoid host keys of a key type being written to console the ssh_key_console_blacklist config key can be used. By default ssh-dss host keys are not written to console.

Internal name: cc_keys_to_console

Module frequency: per instance

Supported distros: all

Config keys:

ssh:
  emit_keys_to_console: false

ssh_fp_console_blacklist: <list of key types>
ssh_key_console_blacklist: <list of key types>

Landscape

Summary: install and configure landscape client

This module installs and configures landscape-client. The landscape client will only be installed if the key landscape is present in config. Landscape client configuration is given under the client key under the main landscape config key. The config parameters are not interpreted by cloud-init, but rather are converted into a ConfigObj formatted file and written out to /etc/landscape/client.conf.

The following default client config is provided, but can be overridden:

landscape:
    client:
        log_level: "info"
        url: "https://landscape.canonical.com/message-system"
        ping_url: "http://landscape.canoncial.com/ping"
        data_path: "/var/lib/landscape/client"

Note

see landscape documentation for client config keys

Note

if tags is defined, its contents should be a string delimited with , rather than a list

Internal name: cc_landscape

Module frequency: per instance

Supported distros: ubuntu

Config keys:

landscape:
    client:
        url: "https://landscape.canonical.com/message-system"
        ping_url: "http://landscape.canonical.com/ping"
        data_path: "/var/lib/landscape/client"
        http_proxy: "http://my.proxy.com/foobar"
        https_proxy: "https://my.proxy.com/foobar"
        tags: "server,cloud"
        computer_title: "footitle"
        registration_key: "fookey"
        account_name: "fooaccount"

Locale

Summary: Set system locale

Configure the system locale and apply it system wide. By default use the locale specified by the datasource.

Internal name: cc_locale

Module frequency: once-per-instance

Supported distros: all

Config schema:

locale: (string) The locale to set as the system’s locale (e.g. ar_PS)

locale_configfile: (string) The file in which to write the locale configuration (defaults to the distro’s default location)

Examples:

# Set the locale to ar_AE
locale: ar_AE

# --- Example2 ---
# Set the locale to fr_CA in /etc/alternate_path/locale
locale: fr_CA
locale_configfile: /etc/alternate_path/locale

LXD

Summary: configure lxd with lxd init and optionally lxd-bridge

This module configures lxd with user specified options using lxd init. If lxd is not present on the system but lxd configuration is provided, then lxd will be installed. If the selected storage backend is zfs, then zfs will be installed if missing. If network bridge configuration is provided, then lxd-bridge will be configured accordingly.

Internal name: cc_lxd

Module frequency: per instance

Supported distros: ubuntu

Config keys:

lxd:
    init:
        network_address: <ip addr>
        network_port: <port>
        storage_backend: <zfs/dir>
        storage_create_device: <dev>
        storage_create_loop: <size>
        storage_pool: <name>
        trust_password: <password>
    bridge:
        mode: <new, existing or none>
        name: <name>
        ipv4_address: <ip addr>
        ipv4_netmask: <cidr>
        ipv4_dhcp_first: <ip addr>
        ipv4_dhcp_last: <ip addr>
        ipv4_dhcp_leases: <size>
        ipv4_nat: <bool>
        ipv6_address: <ip addr>
        ipv6_netmask: <cidr>
        ipv6_nat: <bool>
        domain: <domain>

Mcollective

Summary: install, configure and start mcollective

This module installs, configures and starts mcollective. If the mcollective key is present in config, then mcollective will be installed and started.

Configuration for mcollective can be specified in the conf key under mcollective. Each config value consists of a key value pair and will be written to /etc/mcollective/server.cfg. The public-cert and private-cert keys, if present in conf may be used to specify the public and private certificates for mcollective. Their values will be written to /etc/mcollective/ssl/server-public.pem and /etc/mcollective/ssl/server-private.pem.

Note

The ec2 metadata service is readable by non-root users. If security is a concern, use include-once and ssl urls.

Internal name: cc_mcollective

Module frequency: per instance

Supported distros: all

Config keys:

mcollective:
    conf:
        <key>: <value>
        public-cert: |
            -------BEGIN CERTIFICATE--------
            <cert data>
            -------END CERTIFICATE--------
        private-cert: |
            -------BEGIN CERTIFICATE--------
            <cert data>
            -------END CERTIFICATE--------

Migrator

Summary: migrate old versions of cloud-init data to new

This module handles moving old versions of cloud-init data to newer ones. Currently, it only handles renaming cloud-init’s per-frequency semaphore files to canonicalized name and renaming legacy semaphore names to newer ones. This module is enabled by default, but can be disabled by specifying migrate: false in config.

Internal name: cc_migrator

Module frequency: always

Supported distros: all

Config keys:

migrate: <true/false>

Mounts

Summary: configure mount points and swap files

This module can add or remove mountpoints from /etc/fstab as well as configure swap. The mounts config key takes a list of fstab entries to add. Each entry is specified as a list of [ fs_spec, fs_file, fs_vfstype, fs_mntops, fs-freq, fs_passno ]. For more information on these options, consult the manual for /etc/fstab. When specifying the fs_spec, if the device name starts with one of xvd, sd, hd, or vd, the leading /dev may be omitted.

In order to remove a previously listed mount, an entry can be added to the mounts list containing fs_spec for the device to be removed but no mountpoint (i.e. [ sda1 ] or [ sda1, null ]).

The mount_default_fields config key allows default options to be specified for the values in a mounts entry that are not specified, aside from the fs_spec and the fs_file. If specified, this must be a list containing 6 values. It defaults to:

mount_default_fields: [none, none, "auto", "defaults,nobootwait", "0", "2"]

On a systemd booted system that default is the mostly equivalent:

mount_default_fields: [none, none, "auto",
   "defaults,nofail,x-systemd.requires=cloud-init.service", "0", "2"]

Note that nobootwait is an upstart specific boot option that somewhat equates to the more standard nofail.

Swap files can be configured by setting the path to the swap file to create with filename, the size of the swap file with size maximum size of the swap file if using an size: auto with maxsize. By default no swap file is created.

Internal name: cc_mounts

Module frequency: per instance

Supported distros: all

Config keys:

mounts:
    - [ /dev/ephemeral0, /mnt, auto, "defaults,noexec" ]
    - [ sdc, /opt/data ]
    - [ xvdh, /opt/data, "auto", "defaults,nofail", "0", "0" ]
mount_default_fields: [None, None, "auto", "defaults,nofail", "0", "2"]
swap:
    filename: <file>
    size: <"auto"/size in bytes>
    maxsize: <size in bytes>

NTP

Summary: enable and configure ntp

Handle ntp configuration. If ntp is not installed on the system and ntp configuration is specified, ntp will be installed. If there is a default ntp config file in the image or one is present in the distro’s ntp package, it will be copied to a file with .dist appended to the filename before any changes are made. A list of ntp pools and ntp servers can be provided under the ntp config key. If no ntp servers or pools are provided, 4 pools will be used in the format {0-3}.{distro}.pool.ntp.org.

Internal name: cc_ntp

Module frequency: once-per-instance

Supported distros: almalinux, alpine, centos, cloudlinux, debian, eurolinux, fedora, miraclelinux, openEuler, opensuse, photon, rhel, rocky, sles, ubuntu, virtuozzo

Config schema:

ntp: (object/null)

pools: (array of string) List of ntp pools. If both pools and servers are empty, 4 default pool servers will be provided of the format {0-3}.{distro}.pool.ntp.org. NOTE: for Alpine Linux when using the Busybox NTP client this setting will be ignored due to the limited functionality of Busybox’s ntpd.

servers: (array of string) List of ntp servers. If both pools and servers are empty, 4 default pool servers will be provided with the format {0-3}.{distro}.pool.ntp.org.

ntp_client: (string) Name of an NTP client to use to configure system NTP. When unprovided or ‘auto’ the default client preferred by the distribution will be used. The following built-in client names can be used to override existing configuration defaults: chrony, ntp, ntpdate, systemd-timesyncd.

enabled: (boolean) Attempt to enable ntp clients if set to True. If set to False, ntp client will not be configured or installed

config: (object) Configuration settings or overrides for the ntp_client specified.

confpath: (string) The path to where the ntp_client configuration is written.

check_exe: (string) The executable name for the ntp_client. For example, ntp service check_exe is ‘ntpd’ because it runs the ntpd binary.

packages: (array of string) List of packages needed to be installed for the selected ntp_client.

service_name: (string) The systemd or sysvinit service name used to start and stop the ntp_client service.

template: (string) Inline template allowing users to define their own ntp_client configuration template. The value must start with ‘## template:jinja’ to enable use of templating support.

Examples:

# Override ntp with chrony configuration on Ubuntu
ntp:
  enabled: true
  ntp_client: chrony  # Uses cloud-init default chrony configuration

# --- Example2 ---
# Provide a custom ntp client configuration
ntp:
  enabled: true
  ntp_client: myntpclient
  config:
     confpath: /etc/myntpclient/myntpclient.conf
     check_exe: myntpclientd
     packages:
       - myntpclient
     service_name: myntpclient
     template: |
         ## template:jinja
         # My NTP Client config
         {% if pools -%}# pools{% endif %}
         {% for pool in pools -%}
         pool {{pool}} iburst
         {% endfor %}
         {%- if servers %}# servers
         {% endif %}
         {% for server in servers -%}
         server {{server}} iburst
         {% endfor %}
  pools: [0.int.pool.ntp.org, 1.int.pool.ntp.org, ntp.myorg.org]
  servers:
    - ntp.server.local
    - ntp.ubuntu.com
    - 192.168.23.2

Package Update Upgrade Install

Summary: Update, upgrade, and install packages

This module allows packages to be updated, upgraded or installed during boot. If any packages are to be installed or an upgrade is to be performed then the package cache will be updated first. If a package installation or upgrade requires a reboot, then a reboot can be performed if package_reboot_if_required is specified.

Internal name: cc_package_update_upgrade_install

Module frequency: once-per-instance

Supported distros: all

Config schema:

packages: (array of (array of string)/(string)) A list of packages to install. Each entry in the list can be either a package name or a list with two entries, the first being the package name and the second being the specific package version to install.

package_update: (boolean) Set true to update packages. Happens before upgrade or install. Default: false

package_upgrade: (boolean) Set true to upgrade packages. Happens before install. Default: false

package_reboot_if_required: (boolean) Set true to reboot the system if required by presence of /var/run/reboot-required. Default: false

apt_update: (boolean) DEPRECATED. Use package_update. Default: false

apt_upgrade: (boolean) DEPRECATED. Use package_upgrade. Default: false

apt_reboot_if_required: (boolean) DEPRECATED. Use package_reboot_if_required. Default: false

Examples:

packages:
  - pwgen
  - pastebinit
  - [libpython3.8, 3.8.10-0ubuntu1~20.04.2]
package_update: true
package_upgrade: true
package_reboot_if_required: true

Phone Home

Summary: Post data to url

This module can be used to post data to a remote host after boot is complete. If the post url contains the string $INSTANCE_ID it will be replaced with the id of the current instance. Either all data can be posted or a list of keys to post. Available keys are:

• pub_key_dsa

• pub_key_rsa

• pub_key_ecdsa

• pub_key_ed25519

• instance_id

• hostname

• fdqn

Data is sent as x-www-form-urlencoded arguments.

Example HTTP POST:

POST / HTTP/1.1
Content-Length: 1337
User-Agent: Cloud-Init/21.4
Accept-Encoding: gzip, deflate
Accept: */*
Content-Type: application/x-www-form-urlencoded

pub_key_dsa=dsa_contents&pub_key_rsa=rsa_contents&pub_key_ecdsa=ecdsa_contents&pub_key_ed25519=ed25519_contents&instance_id=i-87018aed&hostname=myhost&fqdn=myhost.internal

Internal name: cc_phone_home

Module frequency: once-per-instance

Supported distros: all

Config schema:

phone_home: (object)

url: (string) The URL to send the phone home data to.

post: (array) A list of keys to post or all. Default: all

tries: (integer) The number of times to try sending the phone home data. Default: 10

Examples:

phone_home:
    url: http://example.com/$INSTANCE_ID/
    post: all

# --- Example2 ---
phone_home:
    url: http://example.com/$INSTANCE_ID/
    post:
        - pub_key_dsa
        - pub_key_rsa
        - pub_key_ecdsa
        - pub_key_ed25519
        - instance_id
        - hostname
        - fqdn
    tries: 5

Power State Change

Summary: Change power state

This module handles shutdown/reboot after all config modules have been run. By default it will take no action, and the system will keep running unless a package installation/upgrade requires a system reboot (e.g. installing a new kernel) and package_reboot_if_required is true.

Using this module ensures that cloud-init is entirely finished with modules that would be executed.

An example to distinguish delay from timeout:

If you delay 5 (5 minutes) and have a timeout of 120 (2 minutes), then the max time until shutdown will be 7 minutes, though it could be as soon as 5 minutes. Cloud-init will invoke ‘shutdown +5’ after the process finishes, or when ‘timeout’ seconds have elapsed.

Note

With Alpine Linux any message value specified is ignored as Alpine’s halt, poweroff, and reboot commands do not support broadcasting a message.

Internal name: cc_power_state_change

Module frequency: once-per-instance

Supported distros: all

Config schema:

power_state: (object)

delay: (integer/string) Time in minutes to delay after cloud-init has finished. Can be now or an integer specifying the number of minutes to delay. Default: now

mode: (string) Must be one of poweroff, halt, or reboot.

message: (string) Optional message to display to the user when the system is powering off or rebooting.

timeout: (integer) Time in seconds to wait for the cloud-init process to finish before executing shutdown. Default: 30

condition: (string/boolean/array) Apply state change only if condition is met. May be boolean true (always met), false (never met), or a command string or list to be executed. For command formatting, see the documentation for cc_runcmd. If exit code is 0, condition is met, otherwise not. Default: true

Examples:

power_state:
    delay: now
    mode: poweroff
    message: Powering off
    timeout: 2
    condition: true

# --- Example2 ---
power_state:
    delay: 30
    mode: reboot
    message: Rebooting machine
    condition: test -f /var/tmp/reboot_me

Puppet

Summary: Install, configure and start puppet

This module handles puppet installation and configuration. If the puppet key does not exist in global configuration, no action will be taken. If a config entry for puppet is present, then by default the latest version of puppet will be installed. If the puppet config key exists in the config archive, this module will attempt to start puppet even if no installation was performed.

The module also provides keys for configuring the new puppet 4 paths and installing the puppet package from the puppetlabs repositories: https://docs.puppet.com/puppet/4.2/reference/whered_it_go.html The keys are package_name, conf_file, ssl_dir and csr_attributes_path. If unset, their values will default to ones that work with puppet 3.x and with distributions that ship modified puppet 4.x that uses the old paths.

Internal name: cc_puppet

Module frequency: once-per-instance

Supported distros: all

Config schema:

puppet: (object)

install: (boolean) Whether or not to install puppet. Setting to false will result in an error if puppet is not already present on the system. Default: true

version: (string) Optional version to pass to the installer script or package manager. If unset, the latest version from the repos will be installed.

install_type: (string) Valid values are packages and aio. Agent packages from the puppetlabs repositories can be installed by setting aio. Based on this setting, the default config/SSL/CSR paths will be adjusted accordingly. Default: packages

collection: (string) Puppet collection to install if install_type is aio. This can be set to one of puppet (rolling release), puppet6, puppet7 (or their nightly counterparts) in order to install specific release streams.

aio_install_url: (string) If install_type is aio, change the url of the install script.

cleanup: (boolean) Whether to remove the puppetlabs repo after installation if install_type is aio Default: true

conf_file: (string) The path to the puppet config file. Default depends on install_type

ssl_dir: (string) The path to the puppet SSL directory. Default depends on install_type

csr_attributes_path: (string) The path to the puppet csr attributes file. Default depends on install_type

package_name: (string) Name of the package to install if install_type is packages. Default: puppet

exec: (boolean) Whether or not to run puppet after configuration finishes. A single manual run can be triggered by setting exec to true, and additional arguments can be passed to puppet agent via the exec_args key (by default the agent will execute with the --test flag). Default: false

exec_args: (array of string) A list of arguments to pass to ‘puppet agent’ if ‘exec’ is true Default: ['--test']

start_service: (boolean) By default, the puppet service will be automatically enabled after installation and set to automatically start on boot. To override this in favor of manual puppet execution set start_service to false

conf: (object) Every key present in the conf object will be added to puppet.conf. As such, section names should be one of: main, server, agent or user and keys should be valid puppet configuration options. The configuration is specified as a dictionary containing high-level <section> keys and lists of <key>=<value> pairs within each section. The certname key supports string substitutions for %i and %f, corresponding to the instance id and fqdn of the machine respectively.

ca_cert is a special case. It won’t be added to puppet.conf. It holds the puppetserver certificate in pem format. It should be a multi-line string (using the | yaml notation for multi-line strings).

main: (object)

server: (object)

agent: (object)

user: (object)

ca_cert: (string)

csr_attributes: (object) create a csr_attributes.yaml file for CSR attributes and certificate extension requests. See https://puppet.com/docs/puppet/latest/config_file_csr_attributes.html

custom_attributes: (object)

extension_requests: (object)

Examples:

puppet:
    install: true
    version: "7.7.0"
    install_type: "aio"
    collection: "puppet7"
    aio_install_url: 'https://git.io/JBhoQ'
    cleanup: true
    conf_file: "/etc/puppet/puppet.conf"
    ssl_dir: "/var/lib/puppet/ssl"
    csr_attributes_path: "/etc/puppet/csr_attributes.yaml"
    exec: true
    exec_args: ['--test']
    conf:
        agent:
            server: "puppetserver.example.org"
            certname: "%i.%f"
        ca_cert: |
            -----BEGIN CERTIFICATE-----
            MIICCTCCAXKgAwIBAgIBATANBgkqhkiG9w0BAQUFADANMQswCQYDVQQDDAJjYTAe
            Fw0xMDAyMTUxNzI5MjFaFw0xNTAyMTQxNzI5MjFaMA0xCzAJBgNVBAMMAmNhMIGf
            MA0GCSqGSIb3DQEBAQUAA4GNADCBiQKBgQCu7Q40sm47/E1Pf+r8AYb/V/FWGPgc
            b014OmNoX7dgCxTDvps/h8Vw555PdAFsW5+QhsGr31IJNI3kSYprFQcYf7A8tNWu
            1MASW2CfaEiOEi9F1R3R4Qlz4ix+iNoHiUDTjazw/tZwEdxaQXQVLwgTGRwVa+aA
            qbutJKi93MILLwIDAQABo3kwdzA4BglghkgBhvhCAQ0EKxYpUHVwcGV0IFJ1Ynkv
            T3BlblNTTCBHZW5lcmF0ZWQgQ2VydGlmaWNhdGUwDwYDVR0TAQH/BAUwAwEB/zAd
            BgNVHQ4EFgQUu4+jHB+GYE5Vxo+ol1OAhevspjAwCwYDVR0PBAQDAgEGMA0GCSqG
            SIb3DQEBBQUAA4GBAH/rxlUIjwNb3n7TXJcDJ6MMHUlwjr03BDJXKb34Ulndkpaf
            +GAlzPXWa7bO908M9I8RnPfvtKnteLbvgTK+h+zX1XCty+S2EQWk29i2AdoqOTxb
            hppiGMp0tT5Havu4aceCXiy2crVcudj3NFciy8X66SoECemW9UYDCb9T5D0d
            -----END CERTIFICATE-----
    csr_attributes:
        custom_attributes:
            1.2.840.113549.1.9.7: 342thbjkt82094y0uthhor289jnqthpc2290
        extension_requests:
            pp_uuid: ED803750-E3C7-44F5-BB08-41A04433FE2E
            pp_image_name: my_ami_image
            pp_preshared_key: 342thbjkt82094y0uthhor289jnqthpc2290

# --- Example2 ---
puppet:
    install_type: "packages"
    package_name: "puppet"
    exec: false

Resizefs

Summary: Resize filesystem

Resize a filesystem to use all avaliable space on partition. This module is useful along with cc_growpart and will ensure that if the root partition has been resized the root filesystem will be resized along with it. By default, cc_resizefs will resize the root partition and will block the boot process while the resize command is running. Optionally, the resize operation can be performed in the background while cloud-init continues running modules. This can be enabled by setting resize_rootfs to noblock. This module can be disabled altogether by setting resize_rootfs to false.

Internal name: cc_resizefs

Module frequency: always

Supported distros: all

Config schema:

resize_rootfs: (true/false/noblock) Whether to resize the root partition. noblock will resize in the background. Default: true

Examples:

resize_rootfs: false  # disable root filesystem resize operation
# --- Example2 ---
resize_rootfs: noblock  # runs resize operation in the background

Resolv Conf

Summary: Configure resolv.conf

This module is intended to manage resolv.conf in environments where early configuration of resolv.conf is necessary for further bootstrapping and/or where configuration management such as puppet or chef own DNS configuration. As Debian/Ubuntu will, by default, utilize resolvconf, and similarly Red Hat will use sysconfig, this module is likely to be of little use unless those are configured correctly.

When using a Config Drive and a RHEL-like system, resolv.conf will also be managed automatically due to the available information provided for DNS servers in the Networking Config Version 2 format. For those that with to have different settings, use this module.

In order for the resolv_conf section to be applied, manage_resolv_conf must be set true.

Note

For Red Hat with sysconfig, be sure to set PEERDNS=no for all DHCP enabled NICs.

Note

And, in Ubuntu/Debian it is recommended that DNS be configured via the standard /etc/network/interfaces configuration file.

Internal name: cc_resolv_conf

Module frequency: once-per-instance

Supported distros: alpine, fedora, opensuse, photon, rhel, sles

Config schema:

manage_resolv_conf: (boolean) Whether to manage the resolv.conf file. resolv_conf block will be ignored unless this is set to true. Default: false

resolv_conf: (object)

nameservers: (array) A list of nameservers to use to be added as nameserver lines

searchdomains: (array) A list of domains to be added search line

domain: (string) The domain to be added as domain line

sortlist: (array) A list of IP addresses to be added to sortlist line

options: (object) Key/value pairs of options to go under options heading. A unary option should be specified as true

Examples:

manage_resolv_conf: true
resolv_conf:
  nameservers:
    - 8.8.8.8
    - 8.8.4.4
  searchdomains:
    - foo.example.com
    - bar.example.com
  domain: example.com
  sortlist:
    - 10.0.0.1/255
    - 10.0.0.2
  options:
    rotate: true
    timeout: 1

Red Hat Subscription

Summary: Register Red Hat Enterprise Linux based system

Register a Red Hat system either by username and password or activation and org. Following a sucessful registration, you can:

• auto-attach subscriptions

• set the service level

• add subscriptions based on pool id

• enable/disable yum repositories based on repo id

• alter the rhsm_baseurl and server-hostname in /etc/rhsm/rhs.conf.

Internal name: cc_rh_subscription

Module frequency: once-per-instance

Supported distros: fedora, rhel

Config schema:

rh_subscription: (object)

username: (string) The username to use. Must be used with password. Should not be used with activation-key or org

password: (string) The password to use. Must be used with username. Should not be used with activation-key or org

activation-key: (string) The activation key to use. Must be used with org. Should not be used with username or password

org: (integer) The organization number to use. Must be used with activation-key. Should not be used with username or password

auto-attach: (boolean) Whether to attach subscriptions automatically

service-level: (string) The service level to use when subscribing to RH repositories. auto-attach must be true for this to be used

add-pool: (array of string) A list of pools ids add to the subscription

enable-repo: (array of string) A list of repositories to enable

disable-repo: (array of string) A list of repositories to disable

rhsm-baseurl: (string) Sets the baseurl in /etc/rhsm/rhsm.conf

server-hostname: (string) Sets the serverurl in /etc/rhsm/rhsm.conf

Examples:

rh_subscription:
    username: joe@foo.bar
    ## Quote your password if it has symbols to be safe
    password: '1234abcd'

# --- Example2 ---
rh_subscription:
    activation-key: foobar
    org: 12345

# --- Example3 ---
rh_subscription:
    activation-key: foobar
    org: 12345
    auto-attach: true
    service-level: self-support
    add-pool:
      - 1a1a1a1a1a1a1a1a1a1a1a1a1a1a1a1a
      - 2b2b2b2b2b2b2b2b2b2b2b2b2b2b2b2b
    enable-repo:
      - repo-id-to-enable
      - other-repo-id-to-enable
    disable-repo:
      - repo-id-to-disable
      - other-repo-id-to-disable
    # Alter the baseurl in /etc/rhsm/rhsm.conf
    rhsm-baseurl: http://url
    # Alter the server hostname in /etc/rhsm/rhsm.conf
    server-hostname: foo.bar.com

RightScale Userdata

Summary: Support rightscale configuration hooks

This module adds support for RightScale configuration hooks to cloud-init. RightScale adds an entry in the format CLOUD_INIT_REMOTE_HOOK=http://... to ec2 user-data. This module checks for this line in the raw userdata and retrieves any scripts linked by the RightScale user data and places them in the user scripts configuration directory, to be run later by cc_scripts_user.

Note

the CLOUD_INIT_REMOTE_HOOK config variable is present in the raw ec2 user data only, not in any cloud-config parts

Raw user data schema:

CLOUD_INIT_REMOTE_HOOK=<url>

Internal name: cc_rightscale_userdata

Module frequency: once-per-instance

Supported distros: all

Config schema:

Rsyslog

Summary: Configure system logging via rsyslog

This module configures remote system logging using rsyslog.

Configuration for remote servers can be specified in configs, but for convenience it can be specified as key value pairs in remotes.

Internal name: cc_rsyslog

Module frequency: once-per-instance

Supported distros: all

Config schema:

rsyslog: (object)

config_dir: (string) The directory where rsyslog configuration files will be written. Default: /etc/rsyslog.d

config_filename: (string) The name of the rsyslog configuration file. Default: 20-cloud-config.conf

configs: (array of (string)/(object)) Each entry in configs is either a string or an object. Each config entry contains a configuration string and a file to write it to. For config entries that are an object, filename sets the target filename and content specifies the config string to write. For config entries that are only a string, the string is used as the config string to write. If the filename to write the config to is not specified, the value of the config_filename key is used. A file with the selected filename will be written inside the directory specified by config_dir.

remotes: (object) Each key is the name for an rsyslog remote entry. Each value holds the contents of the remote config for rsyslog. The config consists of the following parts:

• filter for log messages (defaults to *.*)

• optional leading @ or @@, indicating udp and tcp respectively (defaults to @, for udp)

• ipv4 or ipv6 hostname or address. ipv6 addresses must be in [::1] format, (e.g. @[fd00::1]:514)

• optional port number (defaults to 514)

This module will provide sane defaults for any part of the remote entry that is not specified, so in most cases remote hosts can be specified just using <name>: <address>.

service_reload_command: (array) The command to use to reload the rsyslog service after the config has been updated. If this is set to auto, then an appropriate command for the distro will be used. This is the default behavior. To manually set the command, use a list of command args (e.g. [systemctl, restart, rsyslog]).

Examples:

rsyslog:
    remotes:
        maas: 192.168.1.1
        juju: 10.0.4.1
    service_reload_command: auto

# --- Example2 ---
rsyslog:
    config_dir: /opt/etc/rsyslog.d
    config_filename: 99-late-cloud-config.conf
    configs:
        - "*.* @@192.158.1.1"
        - content: "*.*   @@192.0.2.1:10514"
          filename: 01-example.conf
        - content: |
            *.*   @@syslogd.example.com
    remotes:
        maas: 192.168.1.1
        juju: 10.0.4.1
    service_reload_command: [your, syslog, restart, command]

Runcmd

Summary: Run arbitrary commands

Run arbitrary commands at a rc.local like level with output to the console. Each item can be either a list or a string. If the item is a list, it will be properly quoted. Each item is written to /var/lib/cloud/instance/runcmd to be later interpreted using sh.

Note that the runcmd module only writes the script to be run later. The module that actually runs the script is scripts-user in the Final boot stage.

Note

all commands must be proper yaml, so you have to quote any characters yaml would eat (‘:’ can be problematic)

Note

when writing files, do not use /tmp dir as it races with systemd-tmpfiles-clean LP: #1707222. Use /run/somedir instead.

Internal name: cc_runcmd

Module frequency: once-per-instance

Supported distros: all

Config schema:

runcmd: (array of (array of string)/(string)/(null))

Examples:

runcmd:
    - [ ls, -l, / ]
    - [ sh, -xc, "echo $(date) ': hello world!'" ]
    - [ sh, -c, echo "=========hello world'=========" ]
    - ls -l /root
    - [ wget, "http://example.org", -O, /tmp/index.html ]

Salt Minion

Summary: Setup and run salt minion

This module installs, configures and starts salt minion. If the salt_minion key is present in the config parts, then salt minion will be installed and started. Configuration for salt minion can be specified in the conf key under salt_minion. Any conf values present there will be assigned in /etc/salt/minion. The public and private keys to use for salt minion can be specified with public_key and private_key respectively. Optionally if you have a custom package name, service name or config directory you can specify them with pkg_name, service_name and config_dir.

Salt keys can be manually generated by: salt-key --gen-keys=GEN_KEYS, where GEN_KEYS is the name of the keypair, e.g. ‘minion’. The keypair will be copied to /etc/salt/pki on the minion instance.

Internal name: cc_salt_minion

Module frequency: once-per-instance

Supported distros: all

Config schema:

salt_minion: (object)

pkg_name: (string) Package name to install. Default: salt-minion

service_name: (string) Service name to enable. Default: salt-minion

config_dir: (string) Directory to write config files to. Default: /etc/salt

conf: (object) Configuration to be written to config_dir/minion

grains: (object) Configuration to be written to config_dir/grains

public_key: (string) Public key to be used by the salt minion

private_key: (string) Private key to be used by salt minion

pki_dir: (string) Directory to write key files. Default: config_dir/pki/minion

Examples:

salt_minion:
    pkg_name: salt-minion
    service_name: salt-minion
    config_dir: /etc/salt
    conf:
        master: salt.example.com
    grains:
        role:
            - web
    public_key: |
        ------BEGIN PUBLIC KEY-------
        <key data>
        ------END PUBLIC KEY-------
    private_key: |
        ------BEGIN PRIVATE KEY------
        <key data>
        ------END PRIVATE KEY-------
    pki_dir: /etc/salt/pki/minion

Scripts Per Boot

Summary: run per boot scripts

Any scripts in the scripts/per-boot directory on the datasource will be run every time the system boots. Scripts will be run in alphabetical order. This module does not accept any config keys.

Internal name: cc_scripts_per_boot

Module frequency: always

Supported distros: all

Scripts Per Instance

Summary: run per instance scripts

Any scripts in the scripts/per-instance directory on the datasource will be run when a new instance is first booted. Scripts will be run in alphabetical order. This module does not accept any config keys.

Some cloud platforms change instance-id if a significant change was made to the system. As a result per-instance scripts will run again.

Internal name: cc_scripts_per_instance

Module frequency: per instance

Supported distros: all

Scripts Per Once

Summary: run one time scripts

Any scripts in the scripts/per-once directory on the datasource will be run only once. Changes to the instance will not force a re-run. The only way to re-run these scripts is to run the clean subcommand and reboot. Scripts will be run in alphabetical order. This module does not accept any config keys.

Internal name: cc_scripts_per_once

Module frequency: per once

Supported distros: all

Scripts User

Summary: run user scripts

This module runs all user scripts. User scripts are not specified in the scripts directory in the datasource, but rather are present in the scripts dir in the instance configuration. Any cloud-config parts with a #! will be treated as a script and run. Scripts specified as cloud-config parts will be run in the order they are specified in the configuration. This module does not accept any config keys.

Internal name: cc_scripts_user

Module frequency: per instance

Supported distros: all

Scripts Vendor

Summary: run vendor scripts

Any scripts in the scripts/vendor directory in the datasource will be run when a new instance is first booted. Scripts will be run in alphabetical order. Vendor scripts can be run with an optional prefix specified in the prefix entry under the vendor_data config key.

Internal name: cc_scripts_vendor

Module frequency: per instance

Supported distros: all

Config keys:

vendor_data:
    prefix: <vendor data prefix>

Seed Random

Summary: provide random seed data

Since all cloud instances started from the same image will produce very similar data when they are first booted, as they are all starting with the same seed for the kernel’s entropy keyring. To avoid this, random seed data can be provided to the instance either as a string or by specifying a command to run to generate the data.

Configuration for this module is under the random_seed config key. The file key specifies the path to write the data to, defaulting to /dev/urandom. Data can be passed in directly with data, and may optionally be specified in encoded form, with the encoding specified in encoding.

If the cloud provides its own random seed data, it will be appended to data before it is written to file.

Note

when using a multiline value for data or specifying binary data, be sure to follow yaml syntax and use the | and !binary yaml format specifiers when appropriate

If the command key is specified, the given command will be executed. This will happen after file has been populated. That command’s environment will contain the value of the file key as RANDOM_SEED_FILE. If a command is specified that cannot be run, no error will be reported unless command_required is set to true.

For example, to use pollinate to gather data from a remote entropy server and write it to /dev/urandom, the following could be used:

random_seed:
    file: /dev/urandom
    command: ["pollinate", "--server=http://local.polinate.server"]
    command_required: true

Internal name: cc_seed_random

Module frequency: per instance

Supported distros: all

Config keys:

random_seed:
    file: <file>
    data: <random string>
    encoding: <raw/base64/b64/gzip/gz>
    command: [<cmd name>, <arg1>, <arg2>...]
    command_required: <true/false>

Set Hostname

Summary: set hostname and fqdn

This module handles setting the system hostname and fqdn. If preserve_hostname is set, then the hostname will not be altered.

A hostname and fqdn can be provided by specifying a full domain name under the fqdn key. Alternatively, a hostname can be specified using the hostname key, and the fqdn of the cloud wil be used. If a fqdn specified with the hostname key, it will be handled properly, although it is better to use the fqdn config key. If both fqdn and hostname are set, it is distro dependent whether hostname or fqdn is used, unless the prefer_fqdn_over_hostname option is true and fqdn is set it will force the use of FQDN in all distros, and if false then it will force the hostname use.

This module will run in the init-local stage before networking is configured if the hostname is set by metadata or user data on the local system.

This will occur on datasources like nocloud and ovf where metadata and user data are available locally. This ensures that the desired hostname is applied before any DHCP requests are preformed on these platforms where dynamic DNS is based on initial hostname.

Internal name: cc_set_hostname

Module frequency: always

Supported distros: all

Config keys:

preserve_hostname: <true/false>
prefer_fqdn_over_hostname: <true/false>
fqdn: <fqdn>
hostname: <fqdn/hostname>

Set Passwords

Summary: Set user passwords and enable/disable SSH password authentication

This module consumes three top-level config keys: ssh_pwauth, chpasswd and password.

The ssh_pwauth config key determines whether or not sshd will be configured to accept password authentication. True values will enable password auth, false values will disable password auth, and the literal string unchanged will leave it unchanged. Setting no value will also leave the current setting on-disk unchanged.

The chpasswd config key accepts a dictionary containing either or both of expire and list.

If the list key is provided, it should contain a list of username:password pairs. This can be either a YAML list (of strings), or a multi-line string with one pair per line. Each user will have the corresponding password set. A password can be randomly generated by specifying RANDOM or R as a user’s password. A hashed password, created by a tool like mkpasswd, can be specified; a regex (r'\$(1|2a|2y|5|6)(\$.+){2}') is used to determine if a password value should be treated as a hash.

Note

The users specified must already exist on the system. Users will have been created by the cc_users_groups module at this point.

By default, all users on the system will have their passwords expired (meaning that they will have to be reset the next time the user logs in). To disable this behaviour, set expire under chpasswd to a false value.

If a list of user/password pairs is not specified under chpasswd, then the value of the password config key will be used to set the default user’s password.

Internal name: cc_set_passwords

Module frequency: per instance

Supported distros: all

Config keys:

ssh_pwauth: <yes/no/unchanged>

password: password1
chpasswd:
    expire: <true/false>

chpasswd:
    list: |
        user1:password1
        user2:RANDOM
        user3:password3
        user4:R

##
# or as yaml list
##
chpasswd:
    list:
        - user1:password1
        - user2:RANDOM
        - user3:password3
        - user4:R
        - user4:$6$rL..$ej...

Snap

Summary: Install, configure and manage snapd and snap packages

This module provides a simple configuration namespace in cloud-init to both setup snapd and install snaps.

Note

Both assertions and commands values can be either a dictionary or a list. If these configs are provided as a dictionary, the keys are only used to order the execution of the assertions or commands and the dictionary is merged with any vendor-data snap configuration provided. If a list is provided by the user instead of a dict, any vendor-data snap configuration is ignored.

The assertions configuration option is a dictionary or list of properly-signed snap assertions which will run before any snap commands. They will be added to snapd’s assertion database by invoking snap ack <aggregate_assertion_file>.

Snap commands is a dictionary or list of individual snap commands to run on the target system. These commands can be used to create snap users, install snaps and provide snap configuration.

Note

If ‘side-loading’ private/unpublished snaps on an instance, it is best to create a snap seed directory and seed.yaml manifest in /var/lib/snapd/seed/ which snapd automatically installs on startup.

Development only: The squashfuse_in_container boolean can be set true to install squashfuse package when in a container to enable snap installs. Default is false.

Internal name: cc_snap

Module frequency: once-per-instance

Supported distros: ubuntu

Config schema:

snap: (object)

assertions: (object/array of string)

commands: (object/array of (array of string)/(string))

squashfuse_in_container: (boolean)

Examples:

snap:
    assertions:
      00: |
        signed_assertion_blob_here
      02: |
        signed_assertion_blob_here
    commands:
      00: snap create-user --sudoer --known <snap-user>@mydomain.com
      01: snap install canonical-livepatch
      02: canonical-livepatch enable <AUTH_TOKEN>

# --- Example2 ---
# LXC-based containers require squashfuse before snaps can be installed
snap:
    commands:
        00: apt-get install squashfuse -y
        11: snap install emoj


# --- Example3 ---
# Convenience: the snap command can be omitted when specifying commands
# as a list and 'snap' will automatically be prepended.
# The following commands are equivalent:
snap:
    commands:
        00: ['install', 'vlc']
        01: ['snap', 'install', 'vlc']
        02: snap install vlc
        03: 'snap install vlc'

# --- Example4 ---
# You can use a list of commands
snap:
    commands:
        - ['install', 'vlc']
        - ['snap', 'install', 'vlc']
        - snap install vlc
        - 'snap install vlc'

# --- Example5 ---
# You can use a list of assertions
snap:
    assertions:
        - signed_assertion_blob_here
        - |
            signed_assertion_blob_here

Spacewalk

Summary: install and configure spacewalk

This module installs spacewalk and applies basic configuration. If the spacewalk config key is present spacewalk will be installed. The server to connect to after installation must be provided in the server in spacewalk configuration. A proxy to connect through and a activation key may optionally be specified.

For more information about spacewalk see: https://fedorahosted.org/spacewalk/

Internal name: cc_spacewalk

Module frequency: per instance

Supported distros: redhat, fedora

Config keys:

spacewalk:
   server: <url>
   proxy: <proxy host>
   activation_key: <key>

SSH

Summary: configure SSH and SSH keys (host and authorized)

This module handles most configuration for SSH and both host and authorized SSH keys.

Authorized Keys

Authorized keys are a list of public SSH keys that are allowed to connect to a user account on a system. They are stored in .ssh/authorized_keys in that account’s home directory. Authorized keys for the default user defined in users can be specified using ssh_authorized_keys. Keys should be specified as a list of public keys.

Note

see the cc_set_passwords module documentation to enable/disable SSH password authentication

Root login can be enabled/disabled using the disable_root config key. Root login options can be manually specified with disable_root_opts. If disable_root_opts is specified and contains the string $USER, it will be replaced with the username of the default user. By default, root login is disabled, and root login opts are set to:

no-port-forwarding,no-agent-forwarding,no-X11-forwarding

Supported public key types for the ssh_authorized_keys are:

• dsa

• rsa

• ecdsa

• ed25519

• ecdsa-sha2-nistp256-cert-v01@openssh.com

• ecdsa-sha2-nistp256

• ecdsa-sha2-nistp384-cert-v01@openssh.com

• ecdsa-sha2-nistp384

• ecdsa-sha2-nistp521-cert-v01@openssh.com

• ecdsa-sha2-nistp521

• sk-ecdsa-sha2-nistp256-cert-v01@openssh.com

• sk-ecdsa-sha2-nistp256@openssh.com

• sk-ssh-ed25519-cert-v01@openssh.com

• sk-ssh-ed25519@openssh.com

• ssh-dss-cert-v01@openssh.com

• ssh-dss

• ssh-ed25519-cert-v01@openssh.com

• ssh-ed25519

• ssh-rsa-cert-v01@openssh.com

• ssh-rsa

• ssh-xmss-cert-v01@openssh.com

• ssh-xmss@openssh.com

Note

this list has been filtered out from the supported keytypes of OpenSSH source, where the sigonly keys are removed. Please see ssh_util for more information.

dsa, rsa, ecdsa and ed25519 are added for legacy, as they are valid public keys in some old distros. They can possibly be removed in the future when support for the older distros are dropped

Host Keys

Host keys are for authenticating a specific instance. Many images have default host SSH keys, which can be removed using ssh_deletekeys. This prevents re-use of a private host key from an image on multiple machines. Since removing default host keys is usually the desired behavior this option is enabled by default.

Host keys can be added using the ssh_keys configuration key. The argument to this config key should be a dictionary entries for the public and private keys of each desired key type. Entries in the ssh_keys config dict should have keys in the format <key type>_private, <key type>_public, and, optionally, <key type>_certificate, e.g. rsa_private: <key>, rsa_public: <key>, and rsa_certificate: <key>. See below for supported key types. Not all key types have to be specified, ones left unspecified will not be used. If this config option is used, then no keys will be generated.

When host keys are generated the output of the ssh-keygen command(s) can be displayed on the console using the ssh_quiet_keygen configuration key. This settings defaults to False which displays the keygen output.

Note

when specifying private host keys in cloud-config, care should be taken to ensure that the communication between the data source and the instance is secure

Note

to specify multiline private host keys and certificates, use yaml multiline syntax

If no host keys are specified using ssh_keys, then keys will be generated using ssh-keygen. By default one public/private pair of each supported host key type will be generated. The key types to generate can be specified using the ssh_genkeytypes config flag, which accepts a list of host key types to use. For each host key type for which this module has been instructed to create a keypair, if a key of the same type is already present on the system (i.e. if ssh_deletekeys was false), no key will be generated.

Supported host key types for the ssh_keys and the ssh_genkeytypes config flags are:

• rsa

• dsa

• ecdsa

• ed25519

Internal name: cc_ssh

Module frequency: per instance

Supported distros: all

Config keys:

ssh_deletekeys: <true/false>
ssh_keys:
    rsa_private: |
        -----BEGIN RSA PRIVATE KEY-----
        MIIBxwIBAAJhAKD0YSHy73nUgysO13XsJmd4fHiFyQ+00R7VVu2iV9Qco
        ...
        -----END RSA PRIVATE KEY-----
    rsa_public: ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAGEAoPRhIfLvedSDKw7Xd ...
    rsa_certificate: |
        ssh-rsa-cert-v01@openssh.com AAAAIHNzaC1lZDI1NTE5LWNlcnQt ...
    dsa_private: |
        -----BEGIN DSA PRIVATE KEY-----
        MIIBxwIBAAJhAKD0YSHy73nUgysO13XsJmd4fHiFyQ+00R7VVu2iV9Qco
        ...
        -----END DSA PRIVATE KEY-----
    dsa_public: ssh-dsa AAAAB3NzaC1yc2EAAAABIwAAAGEAoPRhIfLvedSDKw7Xd ...
    dsa_certificate: |
        ssh-dsa-cert-v01@openssh.com AAAAIHNzaC1lZDI1NTE5LWNlcnQt ...

ssh_genkeytypes: <key type>
disable_root: <true/false>
disable_root_opts: <disable root options string>
ssh_authorized_keys:
    - ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAGEA3FSyQwBI6Z+nCSjUU ...
    - ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAQEA3I7VUf2l5gSn5uavROsc5HRDpZ ...
allow_public_ssh_keys: <true/false>
ssh_publish_hostkeys:
    enabled: <true/false> (Defaults to true)
    blacklist: <list of key types> (Defaults to [dsa])
ssh_quiet_keygen: <true/false>

SSH Authkey Fingerprints

Summary: log fingerprints of user SSH keys

Write fingerprints of authorized keys for each user to log. This is enabled by default, but can be disabled using no_ssh_fingerprints. The hash type for the keys can be specified, but defaults to sha256.

Internal name: cc_ssh_authkey_fingerprints

Module frequency: per instance

Supported distros: all

Config keys:

no_ssh_fingerprints: <true/false>
authkey_hash: <hash type>

SSH Import Id

Summary: import SSH id

This module imports SSH keys from either a public keyserver, usually launchpad or github using ssh-import-id. Keys are referenced by the username they are associated with on the keyserver. The keyserver can be specified by prepending either lp: for launchpad or gh: for github to the username.

Internal name: cc_ssh_import_id

Module frequency: per instance

Supported distros: ubuntu, debian

Config keys:

ssh_import_id:
    - user
    - gh:user
    - lp:user

Timezone

Summary: set system timezone

Set the system timezone. If any args are passed to the module then the first will be used for the timezone. Otherwise, the module will attempt to retrieve the timezone from cloud config.

Internal name: cc_timezone

Module frequency: per instance

Supported distros: all

Config keys:

timezone: <timezone>

Ubuntu Advantage

Summary: Configure Ubuntu Advantage support services

Attach machine to an existing Ubuntu Advantage support contract and enable or disable support services such as Livepatch, ESM, FIPS and FIPS Updates. When attaching a machine to Ubuntu Advantage, one can also specify services to enable. When the ‘enable’ list is present, any named service will be enabled and all absent services will remain disabled.

Note that when enabling FIPS or FIPS updates you will need to schedule a reboot to ensure the machine is running the FIPS-compliant kernel. See Power State Change for information on how to configure cloud-init to perform this reboot.

Internal name: cc_ubuntu_advantage

Module frequency: once-per-instance

Supported distros: ubuntu

Config schema:

ubuntu_advantage: (object)

enable: (array of string)

token: (string) A contract token obtained from https://ubuntu.com/advantage.

Examples:

# Attach the machine to an Ubuntu Advantage support contract with a
# UA contract token obtained from https://ubuntu.com/advantage.
ubuntu_advantage:
  token: <ua_contract_token>

# --- Example2 ---
# Attach the machine to an Ubuntu Advantage support contract enabling
# only fips and esm services. Services will only be enabled if
# the environment supports said service. Otherwise warnings will
# be logged for incompatible services specified.
ubuntu-advantage:
  token: <ua_contract_token>
  enable:
  - fips
  - esm

# --- Example3 ---
# Attach the machine to an Ubuntu Advantage support contract and enable
# the FIPS service.  Perform a reboot once cloud-init has
# completed.
power_state:
  mode: reboot
ubuntu-advantage:
  token: <ua_contract_token>
  enable:
  - fips

Ubuntu Drivers

Summary: Interact with third party drivers in Ubuntu.

This module interacts with the ‘ubuntu-drivers’ command to install third party driver packages.

Internal name: cc_ubuntu_drivers

Module frequency: once-per-instance

Supported distros: ubuntu

Config schema:

drivers: (object)

nvidia: (object)

license-accepted: (boolean) Do you accept the NVIDIA driver license?

version: (string) The version of the driver to install (e.g. “390”, “410”). Defaults to the latest version.

Examples:

drivers:
  nvidia:
    license-accepted: true

Update Etc Hosts

Summary: update the hosts file (usually /etc/hosts)

This module will update the contents of the local hosts database (hosts file; usually /etc/hosts) based on the hostname/fqdn specified in config. Management of the hosts file is controlled using manage_etc_hosts. If this is set to false, cloud-init will not manage the hosts file at all. This is the default behavior.

If set to true or template, cloud-init will generate the hosts file using the template located in /etc/cloud/templates/hosts.tmpl. In the /etc/cloud/templates/hosts.tmpl template, the strings $hostname and $fqdn will be replaced with the hostname and fqdn respectively.

If manage_etc_hosts is set to localhost, then cloud-init will not rewrite the hosts file entirely, but rather will ensure that a entry for the fqdn with a distribution dependent ip is present (i.e. ping <hostname> will ping 127.0.0.1 or 127.0.1.1 or other ip).

Note

if manage_etc_hosts is set true or template, the contents of the hosts file will be updated every boot. To make any changes to the hosts file persistent they must be made in /etc/cloud/templates/hosts.tmpl

Note

for instructions on specifying hostname and fqdn, see documentation for cc_set_hostname

Internal name: cc_update_etc_hosts

Module frequency: always

Supported distros: all

Config keys:

manage_etc_hosts: <true/"template"/false/"localhost">
fqdn: <fqdn>
hostname: <fqdn/hostname>

Update Hostname

Summary: update hostname and fqdn

This module will update the system hostname and fqdn. If preserve_hostname is set, then the hostname will not be altered.

Note

for instructions on specifying hostname and fqdn, see documentation for cc_set_hostname

Internal name: cc_update_hostname

Module frequency: always

Supported distros: all

Config keys:

preserve_hostname: <true/false>
prefer_fqdn_over_hostname: <true/false>
fqdn: <fqdn>
hostname: <fqdn/hostname>

Users and Groups

Summary: configure users and groups

This module configures users and groups. For more detailed information on user options, see the Including users and groups config example.

Groups to add to the system can be specified as a list under the groups key. Each entry in the list should either contain a the group name as a string, or a dictionary with the group name as the key and a list of users who should be members of the group as the value. Note: Groups are added before users, so any users in a group list must already exist on the system.

The users config key takes a list of users to configure. The first entry in this list is used as the default user for the system. To preserve the standard default user for the distro, the string default may be used as the first entry of the users list. Each entry in the users list, other than a default entry, should be a dictionary of options for the user. Supported config keys for an entry in users are as follows:

• name: The user’s login name

• expiredate: Optional. Date on which the user’s account will be disabled. Default: none

• gecos: Optional. Comment about the user, usually a comma-separated string of real name and contact information. Default: none

• groups: Optional. Additional groups to add the user to. Default: none

• homedir: Optional. Home dir for user. Default is /home/<username>

• inactive: Optional. Number of days after a password expires until the account is permanently disabled. Default: none

• lock_passwd: Optional. Disable password login. Default: true

• no_create_home: Optional. Do not create home directory. Default: false

• no_log_init: Optional. Do not initialize lastlog and faillog for user. Default: false

• no_user_group: Optional. Do not create group named after user. Default: false

• passwd: Hash of user password

• primary_group: Optional. Primary group for user. Default to new group named after user.

• selinux_user: Optional. SELinux user for user’s login. Default to default SELinux user.

• shell: Optional. The user’s login shell. The default is to set no shell, which results in a system-specific default being used.

• snapuser: Optional. Specify an email address to create the user as a Snappy user through snap create-user. If an Ubuntu SSO account is associated with the address, username and SSH keys will be requested from there. Default: none

• ssh_authorized_keys: Optional. List of SSH keys to add to user’s authkeys file. Default: none. This key can not be combined with ssh_redirect_user.

• ssh_import_id: Optional. SSH id to import for user. Default: none. This key can not be combined with ssh_redirect_user.

• ssh_redirect_user: Optional. Boolean set to true to disable SSH logins for this user. When specified, all cloud meta-data public SSH keys will be set up in a disabled state for this username. Any SSH login as this username will timeout and prompt with a message to login instead as the configured <default_username> for this instance. Default: false. This key can not be combined with ssh_import_id or ssh_authorized_keys.

• sudo: Optional. Sudo rule to use, list of sudo rules to use or False. Default: none. An absence of sudo key, or a value of none or false will result in no sudo rules being written for the user.

• system: Optional. Create user as system user with no home directory. Default: false

• uid: Optional. The user’s ID. Default: The next available value.

Note

Specifying a hash of a user’s password with passwd is a security risk if the cloud-config can be intercepted. SSH authentication is preferred.

Note

If specifying a sudo rule for a user, ensure that the syntax for the rule is valid, as it is not checked by cloud-init.

Note

Most of these configuration options will not be honored if the user already exists. The following options are the exceptions; they are applied to already-existing users: plain_text_passwd, hashed_passwd, lock_passwd, sudo, ssh_authorized_keys, ssh_redirect_user.

Internal name: cc_users_groups

Module frequency: per instance

Supported distros: all

Config keys:

groups:
    - <group>: [<user>, <user>]
    - <group>

users:
    - default
    # User explicitly omitted from sudo permission; also default behavior.
    - name: <some_restricted_user>
      sudo: false
    - name: <username>
      expiredate: '<date>'
      gecos: <comment>
      groups: <additional groups>
      homedir: <home directory>
      inactive: '<number of days>'
      lock_passwd: <true/false>
      no_create_home: <true/false>
      no_log_init: <true/false>
      no_user_group: <true/false>
      passwd: <password>
      primary_group: <primary group>
      selinux_user: <selinux username>
      shell: <shell path>
      snapuser: <email>
      ssh_redirect_user: <true/false>
      ssh_authorized_keys:
          - <key>
          - <key>
      ssh_import_id: <id>
      sudo: <sudo config>
      system: <true/false>
      uid: <user id>

Write Files

Summary: write arbitrary files

Write out arbitrary content to files, optionally setting permissions. Parent folders in the path are created if absent. Content can be specified in plain text or binary. Data encoded with either base64 or binary gzip data can be specified and will be decoded before being written. For empty file creation, content can be omitted.

Note

if multiline data is provided, care should be taken to ensure that it follows yaml formatting standards. to specify binary data, use the yaml option !!binary

Note

Do not write files under /tmp during boot because of a race with systemd-tmpfiles-clean that can cause temp files to get cleaned during the early boot process. Use /run/somedir instead to avoid race LP:1707222.

Internal name: cc_write_files

Module frequency: once-per-instance

Supported distros: all

Config schema:

write_files: (array of object)

Each item in write_files list supports the following keys:

path: (string) Path of the file to which content is decoded and written

content: (string) Optional content to write to the provided path. When content is present and encoding is not ‘text/plain’, decode the content prior to writing. Default: ‘’

owner: (string) Optional owner:group to chown on the file. Default: root:root

permissions: (string) Optional file permissions to set on path represented as an octal string ‘0###’. Default: ‘0644’

encoding: (string) Optional encoding type of the content. Default is text/plain and no content decoding is performed. Supported encoding types are: gz, gzip, gz+base64, gzip+base64, gz+b64, gzip+b64, b64, base64.

append: (boolean) Whether to append content to existing file if path exists. Default: false.

defer: (boolean) Defer writing the file until ‘final’ stage, after users were created, and packages were installed. Default: False.

Examples:

# Write out base64 encoded content to /etc/sysconfig/selinux
write_files:
- encoding: b64
  content: CiMgVGhpcyBmaWxlIGNvbnRyb2xzIHRoZSBzdGF0ZSBvZiBTRUxpbnV4...
  owner: root:root
  path: /etc/sysconfig/selinux
  permissions: '0644'

# --- Example2 ---
# Appending content to an existing file
write_files:
- content: |
    15 * * * * root ship_logs
  path: /etc/crontab
  append: true

# --- Example3 ---
# Provide gziped binary content
write_files:
- encoding: gzip
  content: !!binary |
      H4sIAIDb/U8C/1NW1E/KzNMvzuBKTc7IV8hIzcnJVyjPL8pJ4QIA6N+MVxsAAAA=
  path: /usr/bin/hello
  permissions: '0755'

# --- Example4 ---
# Create an empty file on the system
write_files:
- path: /root/CLOUD_INIT_WAS_HERE

# --- Example5 ---
# Defer writing the file until after the package (Nginx) is
# installed and its user is created alongside
write_files:
- path: /etc/nginx/conf.d/example.com.conf
  content: |
    server {
        server_name example.com;
        listen 80;
        root /var/www;
        location / {
            try_files $uri $uri/ $uri.html =404;
        }
    }
  owner: 'nginx:nginx'
  permissions: '0640'
  defer: true

Yum Add Repo

Summary: add yum repository configuration to the system

Add yum repository configuration to /etc/yum.repos.d. Configuration files are named based on the dictionary key under the yum_repos they are specified with. If a config file already exists with the same name as a config entry, the config entry will be skipped.

Internal name: cc_yum_add_repo

Module frequency: always

Supported distros: almalinux, centos, cloudlinux, eurolinux, fedora,

miraclelinux, openEuler, photon, rhel, rocky, virtuozzo

Config keys:

yum_repos:
    <repo-name>:
        baseurl: <repo url>
        name: <repo name>
        enabled: <true/false>
        # any repository configuration options (see man yum.conf)

Merging User-Data Sections

Overview

This was implemented because it has been a common feature request that there be a way to specify how cloud-config YAML “dictionaries” provided as user-data are merged together when there are multiple YAML files to merge together (say when performing an #include).

Since previously the merging algorithm was very simple and would only overwrite and not append lists, or strings, and so on it was decided to create a new and improved way to merge dictionaries (and their contained objects) together in a way that is customizable, thus allowing for users who provide cloud-config user-data to determine exactly how their objects will be merged.

For example.

#cloud-config (1)
runcmd:
  - bash1
  - bash2

#cloud-config (2)
runcmd:
  - bash3
  - bash4

The previous way of merging the two objects above would result in a final cloud-config object that contains the following.

#cloud-config (merged)
runcmd:
  - bash3
  - bash4

Typically this is not what users want; instead they would likely prefer:

#cloud-config (merged)
runcmd:
  - bash1
  - bash2
  - bash3
  - bash4

This way makes it easier to combine the various cloud-config objects you have into a more useful list, thus reducing duplication necessary to accomplish the same result with the previous method.

Built-in Mergers

Cloud-init provides merging for the following built-in types:

• Dict

• List

• String

The Dict merger has the following options which control what is done with values contained within the config.

• allow_delete: Existing values not present in the new value can be deleted, defaults to False

• no_replace: Do not replace an existing value if one is already present, enabled by default.

• replace: Overwrite existing values with new ones.

The List merger has the following options which control what is done with the values contained within the config.

• append: Add new value to the end of the list, defaults to False.

• prepend: Add new values to the start of the list, defaults to False.

• no_replace: Do not replace an existing value if one is already present, enabled by default.

• replace: Overwrite existing values with new ones.

The Str merger has the following options which control what is done with the values contained within the config.

• append: Add new value to the end of the string, defaults to False.

Common options for all merge types which control how recursive merging is done on other types.

• recurse_dict: If True merge the new values of the dictionary, defaults to True.

• recurse_list: If True merge the new values of the list, defaults to False.

• recurse_array: Alias for recurse_list.

• recurse_str: If True merge the new values of the string, defaults to False.

Customizability

Because the above merging algorithm may not always be desired (just as the previous merging algorithm was not always the preferred one), the concept of customized merging was introduced through ‘merge classes’.

A merge class is a class definition which provides functions that can be used to merge a given type with another given type.

An example of one of these merging classes is the following:

class Merger(object):
    def __init__(self, merger, opts):
        self._merger = merger
        self._overwrite = 'overwrite' in opts

    # This merging algorithm will attempt to merge with
    # another dictionary, on encountering any other type of object
    # it will not merge with said object, but will instead return
    # the original value
    #
    # On encountering a dictionary, it will create a new dictionary
    # composed of the original and the one to merge with, if 'overwrite'
    # is enabled then keys that exist in the original will be overwritten
    # by keys in the one to merge with (and associated values). Otherwise
    # if not in overwrite mode the 2 conflicting keys themselves will
    # be merged.
    def _on_dict(self, value, merge_with):
        if not isinstance(merge_with, (dict)):
            return value
        merged = dict(value)
        for (k, v) in merge_with.items():
            if k in merged:
                if not self._overwrite:
                    merged[k] = self._merger.merge(merged[k], v)
                else:
                    merged[k] = v
            else:
                merged[k] = v
        return merged

As you can see there is a ‘_on_dict’ method here that will be given a source value and a value to merge with. The result will be the merged object. This code itself is called by another merging class which ‘directs’ the merging to happen by analyzing the types of the objects to merge and attempting to find a know object that will merge that type. I will avoid pasting that here, but it can be found in the mergers/__init__.py file (see LookupMerger and UnknownMerger).

So following the typical cloud-init way of allowing source code to be downloaded and used dynamically, it is possible for users to inject there own merging files to handle specific types of merging as they choose (the basic ones included will handle lists, dicts, and strings). Note how each merge can have options associated with it which affect how the merging is performed, for example a dictionary merger can be told to overwrite instead of attempt to merge, or a string merger can be told to append strings instead of discarding other strings to merge with.

How to activate

There are a few ways to activate the merging algorithms, and to customize them for your own usage.

1. The first way involves the usage of MIME messages in cloud-init to specify multipart documents (this is one way in which multiple cloud-config is joined together into a single cloud-config). Two new headers are looked for, both of which can define the way merging is done (the first header to exist wins). These new headers (in lookup order) are ‘Merge-Type’ and ‘X-Merge-Type’. The value should be a string which will satisfy the new merging format definition (see below for this format).

2. The second way is actually specifying the merge-type in the body of the cloud-config dictionary. There are 2 ways to specify this, either as a string or as a dictionary (see format below). The keys that are looked up for this definition are the following (in order), ‘merge_how’, ‘merge_type’.

String format

The string format that is expected is the following.

classname1(option1,option2)+classname2(option3,option4)....

The class name there will be connected to class names used when looking for the class that can be used to merge and options provided will be given to the class on construction of that class.

For example, the default string that is used when none is provided is the following:

list()+dict()+str()

Dictionary format

A dictionary can be used when it specifies the same information as the string format (i.e. the second option above), for example:

{'merge_how': [{'name': 'list', 'settings': ['append']},
               {'name': 'dict', 'settings': ['no_replace', 'recurse_list']},
               {'name': 'str', 'settings': ['append']}]}

This would be the equivalent format for default string format but in dictionary form instead of string form.

Specifying multiple types and its effect

Now you may be asking yourself, if I specify a merge-type header or dictionary for every cloud-config that I provide, what exactly happens?

The answer is that when merging, a stack of ‘merging classes’ is kept, the first one on that stack is the default merging classes, this set of mergers will be used when the first cloud-config is merged with the initial empty cloud-config dictionary. If the cloud-config that was just merged provided a set of merging classes (via the above formats) then those merging classes will be pushed onto the stack. Now if there is a second cloud-config to be merged then the merging classes from the cloud-config before the first will be used (not the default) and so on. This way a cloud-config can decide how it will merge with a cloud-config dictionary coming after it.

Other uses

In addition to being used for merging user-data sections, the default merging algorithm for merging ‘conf.d’ YAML files (which form an initial YAML config for cloud-init) was also changed to use this mechanism so its full benefits (and customization) can also be used there as well. Other places that used the previous merging are also, similarly, now extensible (metadata merging, for example).

Note, however, that merge algorithms are not used across types of configuration. As was the case before merging was implemented, user-data will overwrite conf.d configuration without merging.

Example cloud-config

A common request is to include multiple runcmd directives in different files and merge all of the commands together. To achieve this, we must modify the default merging to allow for dictionaries to join list values.

The first config

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

runcmd:
  - bash1
  - bash2

The second config

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

runcmd:
  - bash3
  - bash4

Instance Metadata

What is instance data?

Instance data is the collection of all configuration data that cloud-init processes to configure the instance. This configuration typically comes from any number of sources:

• cloud-provided metadata services (aka metadata)

• custom config-drive attached to the instance

• cloud-config seed files in the booted cloud image or distribution

• vendordata provided from files or cloud metadata services

• userdata provided at instance creation

Each cloud provider presents unique configuration metadata in different formats to the instance. Cloud-init provides a cache of any crawled metadata as well as a versioned set of standardized instance data keys which it makes available on all platforms.

Cloud-init produces a simple json object in /run/cloud-init/instance-data.json which represents standardized and versioned representation of the metadata it consumes during initial boot. The intent is to provide the following benefits to users or scripts on any system deployed with cloud-init:

• simple static object to query to obtain a instance’s metadata

• speed: avoid costly network transactions for metadata that is already cached on the filesystem

• reduce need to recrawl metadata services for static metadata that is already cached

• leverage cloud-init’s best practices for crawling cloud-metadata services

• avoid rolling unique metadata crawlers on each cloud platform to get metadata configuration values

Cloud-init stores any instance data processed in the following files:

• /run/cloud-init/instance-data.json: world-readable json containing standardized keys, sensitive keys redacted

• /run/cloud-init/instance-data-sensitive.json: root-readable unredacted json blob

• /var/lib/cloud/instance/user-data.txt: root-readable sensitive raw userdata

• /var/lib/cloud/instance/vendor-data.txt: root-readable sensitive raw vendordata

Cloud-init redacts any security sensitive content from instance-data.json, stores /run/cloud-init/instance-data.json as a world-readable json file. Because user-data and vendor-data can contain passwords both of these files are readonly for root as well. The root user can also read /run/cloud-init/instance-data-sensitive.json which is all instance data from instance-data.json as well as unredacted sensitive content.

Format of instance-data.json

The instance-data.json and instance-data-sensitive.json files are well-formed JSON and record the set of keys and values for any metadata processed by cloud-init. Cloud-init standardizes the format for this content so that it can be generalized across different cloud platforms.

There are three basic top-level keys:

• base64_encoded_keys: A list of forward-slash delimited key paths into the instance-data.json object whose value is base64encoded for json compatibility. Values at these paths should be decoded to get the original value.

• sensitive_keys: A list of forward-slash delimited key paths into the instance-data.json object whose value is considered by the datasource as ‘security sensitive’. Only the keys listed here will be redacted from instance-data.json for non-root users.

• merged_cfg: Merged cloud-init ‘system_config’ from /etc/cloud/cloud.cfg and /etc/cloud/cloud-cfg.d. Values under this key could contain sensitive information such as passwords, so it is included in the sensitive-keys list which is only readable by root.

• ds: Datasource-specific metadata crawled for the specific cloud platform. It should closely represent the structure of the cloud metadata crawled. The structure of content and details provided are entirely cloud-dependent. Mileage will vary depending on what the cloud exposes. The content exposed under the ‘ds’ key is currently experimental and expected to change slightly in the upcoming cloud-init release.

• sys_info: Information about the underlying os, python, architecture and kernel. This represents the data collected by cloudinit.util.system_info.

• v1: Standardized cloud-init metadata keys, these keys are guaranteed to exist on all cloud platforms. They will also retain their current behavior and format and will be carried forward even if cloud-init introduces a new version of standardized keys with v2.

The standardized keys present:

v1._beta_keys

List of standardized keys still in ‘beta’. The format, intent or presence of these keys can change. Do not consider them production-ready.

Example output:

• [subplatform]

v1.cloud_name

Where possible this will indicate the ‘name’ of the cloud the system is running on. This is different than the ‘platform’ item. For example, the cloud name of Amazon Web Services is ‘aws’, while the platform is ‘ec2’.

If determining a specific name is not possible or provided in meta-data, then this filed may contain the same content as ‘platform’.

Example output:

• aws

• openstack

• azure

• configdrive

• nocloud

• ovf

v1.distro, v1.distro_version, v1.distro_release

This shall be the distro name, version and release as determined by cloudinit.util.get_linux_distro.

Example output:

• alpine, 3.12.0, ‘’

• centos, 7.5, core

• debian, 9, stretch

• freebsd, 12.0-release-p10,

• opensuse, 42.3, x86_64

• opensuse-tumbleweed, 20180920, x86_64

• redhat, 7.5, ‘maipo’

• sles, 12.3, x86_64

• ubuntu, 20.04, focal

v1.instance_id

Unique instance_id allocated by the cloud.

Examples output:

• i-<hash>

v1.kernel_release

This shall be the running kernel uname -r

Example output:

• 5.3.0-1010-aws

v1.local_hostname

The internal or local hostname of the system.

Examples output:

• ip-10-41-41-70

• <user-provided-hostname>

v1.machine

This shall be the running cpu machine architecture uname -m

Example output:

• x86_64

• i686

• ppc64le

• s390x

v1.platform

An attempt to identify the cloud platform instance that the system is running on.

Examples output:

• ec2

• openstack

• lxd

• gce

• nocloud

• ovf

v1.subplatform

Additional platform details describing the specific source or type of metadata used. The format of subplatform will be:

<subplatform_type> (<url_file_or_dev_path>)

Examples output:

• metadata (http://168.254.169.254)

• seed-dir (/path/to/seed-dir/)

• config-disk (/dev/cd0)

• configdrive (/dev/sr0)

v1.public_ssh_keys

A list of SSH keys provided to the instance by the datasource metadata.

Examples output:

• [‘ssh-rsa AA…’, …]

v1.python_version

The version of python that is running cloud-init as determined by cloudinit.util.system_info

Example output:

• 3.7.6

v1.region

The physical region/data center in which the instance is deployed.

Examples output:

• us-east-2

v1.availability_zone

The physical availability zone in which the instance is deployed.

Examples output:

• us-east-2b

• nova

• null

Example Output

Below is an example of /run/cloud-init/instance-data-sensitive.json on an EC2 instance:

{
 "_beta_keys": [
  "subplatform"
 ],
 "availability_zone": "us-east-1b",
 "base64_encoded_keys": [],
 "merged_cfg": {
  "_doc": "Merged cloud-init system config from /etc/cloud/cloud.cfg and /etc/cloud/cloud.cfg.d/",
  "_log": [
   "[loggers]\nkeys=root,cloudinit\n\n[handlers]\nkeys=consoleHandler,cloudLogHandler\n\n[formatters]\nkeys=simpleFormatter,arg0Formatter\n\n[logger_root]\nlevel=DEBUG\nhandlers=consoleHandler,cloudLogHandler\n\n[logger_cloudinit]\nlevel=DEBUG\nqualname=cloudinit\nhandlers=\npropagate=1\n\n[handler_consoleHandler]\nclass=StreamHandler\nlevel=WARNING\nformatter=arg0Formatter\nargs=(sys.stderr,)\n\n[formatter_arg0Formatter]\nformat=%(asctime)s - %(filename)s[%(levelname)s]: %(message)s\n\n[formatter_simpleFormatter]\nformat=[CLOUDINIT] %(filename)s[%(levelname)s]: %(message)s\n",
   "[handler_cloudLogHandler]\nclass=FileHandler\nlevel=DEBUG\nformatter=arg0Formatter\nargs=('/var/log/cloud-init.log',)\n",
   "[handler_cloudLogHandler]\nclass=handlers.SysLogHandler\nlevel=DEBUG\nformatter=simpleFormatter\nargs=(\"/dev/log\", handlers.SysLogHandler.LOG_USER)\n"
  ],
  "cloud_config_modules": [
   "emit_upstart",
   "snap",
   "ssh-import-id",
   "locale",
   "set-passwords",
   "grub-dpkg",
   "apt-pipelining",
   "apt-configure",
   "ubuntu-advantage",
   "ntp",
   "timezone",
   "disable-ec2-metadata",
   "runcmd",
   "byobu"
  ],
  "cloud_final_modules": [
   "package-update-upgrade-install",
   "fan",
   "landscape",
   "lxd",
   "ubuntu-drivers",
   "puppet",
   "chef",
   "mcollective",
   "salt-minion",
   "rightscale_userdata",
   "scripts-vendor",
   "scripts-per-once",
   "scripts-per-boot",
   "scripts-per-instance",
   "scripts-user",
   "ssh-authkey-fingerprints",
   "keys-to-console",
   "phone-home",
   "final-message",
   "power-state-change"
  ],
  "cloud_init_modules": [
   "migrator",
   "seed_random",
   "bootcmd",
   "write-files",
   "growpart",
   "resizefs",
   "disk_setup",
   "mounts",
   "set_hostname",
   "update_hostname",
   "update_etc_hosts",
   "ca-certs",
   "rsyslog",
   "users-groups",
   "ssh"
  ],
  "datasource_list": [
   "Ec2",
   "None"
  ],
  "def_log_file": "/var/log/cloud-init.log",
  "disable_root": true,
  "log_cfgs": [
   [
    "[loggers]\nkeys=root,cloudinit\n\n[handlers]\nkeys=consoleHandler,cloudLogHandler\n\n[formatters]\nkeys=simpleFormatter,arg0Formatter\n\n[logger_root]\nlevel=DEBUG\nhandlers=consoleHandler,cloudLogHandler\n\n[logger_cloudinit]\nlevel=DEBUG\nqualname=cloudinit\nhandlers=\npropagate=1\n\n[handler_consoleHandler]\nclass=StreamHandler\nlevel=WARNING\nformatter=arg0Formatter\nargs=(sys.stderr,)\n\n[formatter_arg0Formatter]\nformat=%(asctime)s - %(filename)s[%(levelname)s]: %(message)s\n\n[formatter_simpleFormatter]\nformat=[CLOUDINIT] %(filename)s[%(levelname)s]: %(message)s\n",
    "[handler_cloudLogHandler]\nclass=FileHandler\nlevel=DEBUG\nformatter=arg0Formatter\nargs=('/var/log/cloud-init.log',)\n"
   ]
  ],
  "output": {
   "all": "| tee -a /var/log/cloud-init-output.log"
  },
  "preserve_hostname": false,
  "syslog_fix_perms": [
   "syslog:adm",
   "root:adm",
   "root:wheel",
   "root:root"
  ],
  "users": [
   "default"
  ],
  "vendor_data": {
   "enabled": true,
   "prefix": []
  }
 },
 "cloud_name": "aws",
 "distro": "ubuntu",
 "distro_release": "focal",
 "distro_version": "20.04",
 "ds": {
  "_doc": "EXPERIMENTAL: The structure and format of content scoped under the 'ds' key may change in subsequent releases of cloud-init.",
  "_metadata_api_version": "2016-09-02",
  "dynamic": {
   "instance_identity": {
    "document": {
     "accountId": "329910648901",
     "architecture": "x86_64",
     "availabilityZone": "us-east-1b",
     "billingProducts": null,
     "devpayProductCodes": null,
     "imageId": "ami-02e8aa396f8be3b6d",
     "instanceId": "i-0929128ff2f73a2f1",
     "instanceType": "t2.micro",
     "kernelId": null,
     "marketplaceProductCodes": null,
     "pendingTime": "2020-02-27T20:46:18Z",
     "privateIp": "172.31.81.43",
     "ramdiskId": null,
     "region": "us-east-1",
     "version": "2017-09-30"
    },
    "pkcs7": [
     "MIAGCSqGSIb3DQ...",
     "REDACTED",
     "AhQUgq0iPWqPTVnT96tZE6L1XjjLHQAAAAAAAA=="
    ],
    "rsa2048": [
     "MIAGCSqGSIb...",
     "REDACTED",
     "clYQvuE45xXm7Yreg3QtQbrP//owl1eZHj6s350AAAAAAAA="
    ],
    "signature": [
     "dA+QV+LLCWCRNddnrKleYmh2GvYo+t8urDkdgmDSsPi",
     "REDACTED",
     "kDT4ygyJLFkd3b4qjAs="
    ]
   }
  },
  "meta_data": {
   "ami_id": "ami-02e8aa396f8be3b6d",
   "ami_launch_index": "0",
   "ami_manifest_path": "(unknown)",
   "block_device_mapping": {
    "ami": "/dev/sda1",
    "root": "/dev/sda1"
   },
   "hostname": "ip-172-31-81-43.ec2.internal",
   "instance_action": "none",
   "instance_id": "i-0929128ff2f73a2f1",
   "instance_type": "t2.micro",
   "local_hostname": "ip-172-31-81-43.ec2.internal",
   "local_ipv4": "172.31.81.43",
   "mac": "12:7e:c9:93:29:af",
   "metrics": {
    "vhostmd": "<?xml version=\"1.0\" encoding=\"UTF-8\"?>"
   },
   "network": {
    "interfaces": {
     "macs": {
      "12:7e:c9:93:29:af": {
       "device_number": "0",
       "interface_id": "eni-0c07a0474339b801d",
       "ipv4_associations": {
        "3.89.187.177": "172.31.81.43"
       },
       "local_hostname": "ip-172-31-81-43.ec2.internal",
       "local_ipv4s": "172.31.81.43",
       "mac": "12:7e:c9:93:29:af",
       "owner_id": "329910648901",
       "public_hostname": "ec2-3-89-187-177.compute-1.amazonaws.com",
       "public_ipv4s": "3.89.187.177",
       "security_group_ids": "sg-0100038b68aa79986",
       "security_groups": "launch-wizard-3",
       "subnet_id": "subnet-04e2d12a",
       "subnet_ipv4_cidr_block": "172.31.80.0/20",
       "vpc_id": "vpc-210b4b5b",
       "vpc_ipv4_cidr_block": "172.31.0.0/16",
       "vpc_ipv4_cidr_blocks": "172.31.0.0/16"
      }
     }
    }
   },
   "placement": {
    "availability_zone": "us-east-1b"
   },
   "profile": "default-hvm",
   "public_hostname": "ec2-3-89-187-177.compute-1.amazonaws.com",
   "public_ipv4": "3.89.187.177",
   "reservation_id": "r-0c481643d15766a02",
   "security_groups": "launch-wizard-3",
   "services": {
    "domain": "amazonaws.com",
    "partition": "aws"
   }
  }
 },
 "instance_id": "i-0929128ff2f73a2f1",
 "kernel_release": "5.3.0-1010-aws",
 "local_hostname": "ip-172-31-81-43",
 "machine": "x86_64",
 "platform": "ec2",
 "public_ssh_keys": [],
 "python_version": "3.7.6",
 "region": "us-east-1",
 "sensitive_keys": [],
 "subplatform": "metadata (http://169.254.169.254)",
 "sys_info": {
  "dist": [
   "ubuntu",
   "20.04",
   "focal"
  ],
  "platform": "Linux-5.3.0-1010-aws-x86_64-with-Ubuntu-20.04-focal",
  "python": "3.7.6",
  "release": "5.3.0-1010-aws",
  "system": "Linux",
  "uname": [
   "Linux",
   "ip-172-31-81-43",
   "5.3.0-1010-aws",
   "#11-Ubuntu SMP Thu Jan 16 07:59:32 UTC 2020",
   "x86_64",
   "x86_64"
  ],
  "variant": "ubuntu"
 },
 "system_platform": "Linux-5.3.0-1010-aws-x86_64-with-Ubuntu-20.04-focal",
 "userdata": "#cloud-config\nssh_import_id: [<my-launchpad-id>]\n...",
 "v1": {
  "_beta_keys": [
   "subplatform"
  ],
  "availability_zone": "us-east-1b",
  "cloud_name": "aws",
  "distro": "ubuntu",
  "distro_release": "focal",
  "distro_version": "20.04",
  "instance_id": "i-0929128ff2f73a2f1",
  "kernel": "5.3.0-1010-aws",
  "local_hostname": "ip-172-31-81-43",
  "machine": "x86_64",
  "platform": "ec2",
  "public_ssh_keys": [],
  "python": "3.7.6",
  "region": "us-east-1",
  "subplatform": "metadata (http://169.254.169.254)",
  "system_platform": "Linux-5.3.0-1010-aws-x86_64-with-Ubuntu-20.04-focal",
  "variant": "ubuntu"
 },
 "variant": "ubuntu",
 "vendordata": ""
}

Using instance-data

As of cloud-init v. 18.4, any instance-data can be used in:

• User-data scripts

• Cloud config data

• Command line interface via cloud-init query or cloud-init devel render

This means that any variable present in /run/cloud-init/instance-data-sensitive.json can be used, unless a non-root user is using the command line interface. In the non-root user case, /run/cloud-init/instance-data.json will be used instead.

Many clouds allow users to provide user-data to an instance at the time the instance is launched. Cloud-init supports a number of User-Data Formats.

Both user-data scripts and #cloud-config data support jinja template rendering. When the first line of the provided user-data begins with, ## template: jinja cloud-init will use jinja to render that file. Any instance-data-sensitive.json variables are surfaced as jinja template variables because cloud-config modules are run as ‘root’ user.

Note

cloud-init also provides jinja-safe key aliases for any instance-data.json keys which contain jinja operator characters such as +, -, ., /, etc. Any jinja operator will be replaced with underscores in the jinja-safe key alias. This allows for cloud-init templates to use aliased variable references which allow for jinja’s dot-notation reference such as {{ ds.v1_0.my_safe_key }} instead of {{ ds["v1.0"]["my/safe-key"] }}.

Below are some other examples of using jinja templates in user-data:

• Cloud config calling home with the ec2 public hostname and availability-zone

## template: jinja
#cloud-config
runcmd:
    - echo 'EC2 public hostname allocated to instance: {{
      ds.meta_data.public_hostname }}' > /tmp/instance_metadata
    - echo 'EC2 availability zone: {{ v1.availability_zone }}' >>
      /tmp/instance_metadata
    - curl -X POST -d '{"hostname": "{{ds.meta_data.public_hostname }}",
      "availability-zone": "{{ v1.availability_zone }}"}'
      https://example.com

• Custom user-data script performing different operations based on region

## template: jinja
#!/bin/bash
{% if v1.region == 'us-east-2' -%}
echo 'Installing custom proxies for {{ v1.region }}
sudo apt-get install my-xtra-fast-stack
{%- endif %}
...

One way to easily explore what Jinja variables are available on your machine is to use the cloud-init query --format (-f) commandline option which will render any Jinja syntax you use. Warnings or exceptions will be raised on invalid instance-data keys, paths or invalid syntax.

# List all instance-data keys and values as root user
% sudo cloud-init query --all
{...}

# Introspect nested keys on an object
% cloud-init query -f "{{ds.keys()}}"
dict_keys(['meta_data', '_doc'])

# Test your Jinja rendering syntax on the command-line directly

# Failure to reference valid top-level instance-data key
% cloud-init query -f "{{invalid.instance-data.key}}"
WARNING: Ignoring jinja template for query commandline: 'invalid' is undefined

# Failure to reference valid dot-delimited key path on a known top-level key
% cloud-init query -f "{{v1.not_here}}"
WARNING: Could not render jinja template variables in file 'query commandline': 'not_here'
CI_MISSING_JINJA_VAR/not_here

# Test expected value using valid instance-data key path
% cloud-init query -f "My AMI: {{ds.meta_data.ami_id}}"
My AMI: ami-0fecc35d3c8ba8d60

Note

Trying to reference jinja variables that don’t exist in instance-data will result in warnings in /var/log/cloud-init.log and the following string in your rendered user-data: CI_MISSING_JINJA_VAR/<your_varname>.

Cloud-init also surfaces a command line tool cloud-init query which can assist developers or scripts with obtaining instance metadata easily. See query for more information.

To cut down on keystrokes on the command line, cloud-init also provides top-level key aliases for any standardized v# keys present. The preceding v1 is not required of v1.var_name These aliases will represent the value of the highest versioned standard key. For example, cloud_name value will be v2.cloud_name if both v1 and v2 keys are present in instance-data.json. The query command also publishes userdata and vendordata keys to the root user which will contain the decoded user and vendor data provided to this instance. Non-root users referencing userdata or vendordata keys will see only redacted values.

# List all top-level instance-data keys available
% cloud-init query --list-keys

# Find your EC2 ami-id
% cloud-init query ds.metadata.ami_id

# Format your cloud_name and region using jinja template syntax
% cloud-init query --format 'cloud: {{ v1.cloud_name }} myregion: {{
% v1.region }}'

# Locally test that your template userdata provided to the vm was rendered as
# intended.
% cloud-init query --format "$(sudo cloud-init query userdata)"

# The --format command renders jinja templates, this can also be used
# to develop and test jinja template constructs
% cat > test-templating.yaml <<EOF
  {% for val in ds.meta_data.keys() %}
  - {{ val }}
  {% endfor %}
  EOF
% cloud-init query --format="$( cat test-templating.yaml )"
- instance_id
- dsmode
- local_hostname

Note

To save time designing a user-data template for a specific cloud’s instance-data.json, use the ‘render’ cloud-init command on an instance booted on your favorite cloud. See devel for more information.

Datasources

Datasources are sources of configuration data for cloud-init that typically come from the user (i.e. userdata) or come from the cloud that created the configuration drive (i.e. metadata). Typical userdata would include files, YAML, and shell scripts while typical metadata would include server name, instance id, display name and other cloud specific details.

Since there are multiple ways to provide this data (each cloud solution seems to prefer its own way) internally a datasource abstract class was created to allow for a single way to access the different cloud systems methods to provide this data through the typical usage of subclasses.

Any metadata processed by cloud-init’s datasources is persisted as /run/cloud-init/instance-data.json. Cloud-init provides tooling to quickly introspect some of that data. See Instance Metadata for more information.

Known Sources

The following is a list of documents for each supported datasource:

Alibaba Cloud (AliYun)

The AliYun datasource reads data from Alibaba Cloud ECS. Support is present in cloud-init since 0.7.9.

Metadata Service

The Alibaba Cloud metadata service is available at the well known url http://100.100.100.200/. For more information see Alibaba Cloud ECS on metadata.

Configuration

The following configuration can be set for the datasource in system configuration (in /etc/cloud/cloud.cfg or /etc/cloud/cloud.cfg.d/).

An example configuration with the default values is provided below:

datasource:
  AliYun:
    metadata_urls: ["http://100.100.100.200"]
    timeout: 50
    max_wait: 120

Versions

Like the EC2 metadata service, Alibaba Cloud’s metadata service provides versioned data under specific paths. As of April 2018, there are only 2016-01-01 and latest versions.

It is expected that the dated version will maintain a stable interface but latest may change content at a future date.

Cloud-init uses the 2016-01-01 version.

You can list the versions available to your instance with:

$ curl http://100.100.100.200/
2016-01-01
latest

Metadata

Instance metadata can be queried at http://100.100.100.200/2016-01-01/meta-data

$ curl http://100.100.100.200/2016-01-01/meta-data
dns-conf/
eipv4
hostname
image-id
instance-id
instance/
mac
network-type
network/
ntp-conf/
owner-account-id
private-ipv4
public-keys/
region-id
serial-number
source-address
sub-private-ipv4-list
vpc-cidr-block
vpc-id

Userdata

If provided, user-data will appear at http://100.100.100.200/2016-01-01/user-data. If no user-data is provided, this will return a 404.

$ curl http://100.100.100.200/2016-01-01/user-data
#!/bin/sh
echo "Hello World."

Alt Cloud

The datasource altcloud will be used to pick up user data on RHEVm and vSphere.

RHEVm

For RHEVm v3.0 the userdata is injected into the VM using floppy injection via the RHEVm dashboard “Custom Properties”.

The format of the Custom Properties entry must be:

floppyinject=user-data.txt:<base64 encoded data>

For example to pass a simple bash script:

% cat simple_script.bash
#!/bin/bash
echo "Hello Joe!" >> /tmp/JJV_Joe_out.txt

% base64 < simple_script.bash
IyEvYmluL2Jhc2gKZWNobyAiSGVsbG8gSm9lISIgPj4gL3RtcC9KSlZfSm9lX291dC50eHQK

To pass this example script to cloud-init running in a RHEVm v3.0 VM set the “Custom Properties” when creating the RHEMv v3.0 VM to:

floppyinject=user-data.txt:IyEvYmluL2Jhc2gKZWNobyAiSGVsbG8gSm9lISIgPj4gL3RtcC9KSlZfSm9lX291dC50eHQK

NOTE: The prefix with file name must be: floppyinject=user-data.txt:

It is also possible to launch a RHEVm v3.0 VM and pass optional user data to it using the Delta Cloud.

For more information on Delta Cloud see: http://deltacloud.apache.org

vSphere

For VMWare’s vSphere the userdata is injected into the VM as an ISO via the cdrom. This can be done using the vSphere dashboard by connecting an ISO image to the CD/DVD drive.

To pass this example script to cloud-init running in a vSphere VM set the CD/DVD drive when creating the vSphere VM to point to an ISO on the data store.

Note: The ISO must contain the user data.

For example, to pass the same simple_script.bash to vSphere:

Create the ISO

% mkdir my-iso

NOTE: The file name on the ISO must be: user-data.txt

% cp simple_script.bash my-iso/user-data.txt
% genisoimage -o user-data.iso -r my-iso

Verify the ISO

% sudo mkdir /media/vsphere_iso
% sudo mount -o loop user-data.iso /media/vsphere_iso
% cat /media/vsphere_iso/user-data.txt
% sudo umount /media/vsphere_iso

Then, launch the vSphere VM the ISO user-data.iso attached as a CDROM.

It is also possible to launch a vSphere VM and pass optional user data to it using the Delta Cloud.

For more information on Delta Cloud see: http://deltacloud.apache.org

Amazon EC2

The EC2 datasource is the oldest and most widely used datasource that cloud-init supports. This datasource interacts with a magic ip that is provided to the instance by the cloud provider. Typically this ip is 169.254.169.254 of which at this ip a http server is provided to the instance so that the instance can make calls to get instance userdata and instance metadata.

Metadata is accessible via the following URL:

GET http://169.254.169.254/2009-04-04/meta-data/
ami-id
ami-launch-index
ami-manifest-path
block-device-mapping/
hostname
instance-id
instance-type
local-hostname
local-ipv4
placement/
public-hostname
public-ipv4
public-keys/
reservation-id
security-groups

Userdata is accessible via the following URL:

GET http://169.254.169.254/2009-04-04/user-data
1234,fred,reboot,true | 4512,jimbo, | 173,,,

Note that there are multiple versions of this data provided, cloud-init by default uses 2009-04-04 but newer versions can be supported with relative ease (newer versions have more data exposed, while maintaining backward compatibility with the previous versions). Version 2016-09-02 is required for secondary IP address support.

To see which versions are supported from your cloud provider use the following URL:

GET http://169.254.169.254/
1.0
2007-01-19
2007-03-01
2007-08-29
2007-10-10
2007-12-15
2008-02-01
2008-09-01
2009-04-04
...
latest

Configuration

The following configuration can be set for the datasource in system configuration (in /etc/cloud/cloud.cfg or /etc/cloud/cloud.cfg.d/).

The settings that may be configured are:

• metadata_urls: This list of urls will be searched for an Ec2 metadata service. The first entry that successfully returns a 200 response for <url>/<version>/meta-data/instance-id will be selected. (default: [‘http://169.254.169.254’, ‘http://instance-data:8773’]).

• max_wait: the maximum amount of clock time in seconds that should be spent searching metadata_urls. A value less than zero will result in only one request being made, to the first in the list. (default: 120)

• timeout: the timeout value provided to urlopen for each individual http request. This is used both when selecting a metadata_url and when crawling the metadata service. (default: 50)

• apply_full_imds_network_config: Boolean (default: True) to allow cloud-init to configure any secondary NICs and secondary IPs described by the metadata service. All network interfaces are configured with DHCP (v4) to obtain an primary IPv4 address and route. Interfaces which have a non-empty ‘ipv6s’ list will also enable DHCPv6 to obtain a primary IPv6 address and route. The DHCP response (v4 and v6) return an IP that matches the first element of local-ipv4s and ipv6s lists respectively. All additional values (secondary addresses) in the static ip lists will be added to interface.

An example configuration with the default values is provided below:

datasource:
  Ec2:
    metadata_urls: ["http://169.254.169.254:80", "http://instance-data:8773"]
    max_wait: 120
    timeout: 50
    apply_full_imds_network_config: true

Notes

• There are 2 types of EC2 instances network-wise: VPC ones (Virtual Private Cloud) and Classic ones (also known as non-VPC). One major difference between them is that Classic instances have their MAC address changed on stop/restart operations, so cloud-init will recreate the network config file for EC2 Classic instances every boot. On VPC instances this file is generated only in the first boot of the instance. The check for the instance type is performed by is_classic_instance() method.

• For EC2 instances with multiple network interfaces (NICs) attached, dhcp4 will be enabled to obtain the primary private IPv4 address of those NICs. Wherever dhcp4 or dhcp6 is enabled for a NIC, a dhcp route-metric will be added with the value of <device-number + 1> * 100 to ensure dhcp routes on the primary NIC are preferred to any secondary NICs. For example: the primary NIC will have a DHCP route-metric of 100, the next NIC will be 200.

Azure

This datasource finds metadata and user-data from the Azure cloud platform.

The Azure cloud platform provides initial data to an instance via an attached CD formatted in UDF. That CD contains a ‘ovf-env.xml’ file that provides some information. Additional information is obtained via interaction with the “endpoint”.

IMDS

Azure provides the instance metadata service (IMDS) which is a REST service on 169.254.169.254 providing additional configuration information to the instance. Cloud-init uses the IMDS for:

• network configuration for the instance which is applied per boot

• a preprovisioing gate which blocks instance configuration until Azure fabric is ready to provision

• retrieving SSH public keys. Cloud-init will first try to utilize SSH keys returned from IMDS, and if they are not provided from IMDS then it will fallback to using the OVF file provided from the CD-ROM. There is a large performance benefit to using IMDS for SSH key retrieval, but in order to support environments where IMDS is not available then we must continue to all for keys from OVF

Configuration

The following configuration can be set for the datasource in system configuration (in /etc/cloud/cloud.cfg or /etc/cloud/cloud.cfg.d/).

The settings that may be configured are:

• apply_network_config: Boolean set to True to use network configuration described by Azure’s IMDS endpoint instead of fallback network config of dhcp on eth0. Default is True. For Ubuntu 16.04 or earlier, default is False.

• data_dir: Path used to read metadata files and write crawled data.

• disk_aliases: A dictionary defining which device paths should be interpreted as ephemeral images. See cc_disk_setup module for more info.

Configuration for the datasource can also be read from a dscfg entry in the LinuxProvisioningConfigurationSet. Content in dscfg node is expected to be base64 encoded yaml content, and it will be merged into the ‘datasource: Azure’ entry.

An example configuration with the default values is provided below:

datasource:
  Azure:
    apply_network_config: true
    data_dir: /var/lib/waagent
    disk_aliases:
      ephemeral0: /dev/disk/cloud/azure_resource

Userdata

Userdata is provided to cloud-init inside the ovf-env.xml file. Cloud-init expects that user-data will be provided as base64 encoded value inside the text child of a element named UserData or CustomData which is a direct child of the LinuxProvisioningConfigurationSet (a sibling to UserName) If both UserData and CustomData are provided behavior is undefined on which will be selected.

In the example below, user-data provided is ‘this is my userdata’

Example:

<wa:ProvisioningSection>
 <wa:Version>1.0</wa:Version>
 <LinuxProvisioningConfigurationSet
    xmlns="http://schemas.microsoft.com/windowsazure"
    xmlns:i="http://www.w3.org/2001/XMLSchema-instance">
  <ConfigurationSetType>LinuxProvisioningConfiguration</ConfigurationSetType>
  <HostName>myHost</HostName>
  <UserName>myuser</UserName>
  <UserPassword/>
  <CustomData>dGhpcyBpcyBteSB1c2VyZGF0YQ===</CustomData>
  <dscfg>eyJhZ2VudF9jb21tYW5kIjogWyJzdGFydCIsICJ3YWxpbnV4YWdlbnQiXX0=</dscfg>
  <DisableSshPasswordAuthentication>true</DisableSshPasswordAuthentication>
  <SSH>
   <PublicKeys>
    <PublicKey>
     <Fingerprint>6BE7A7C3C8A8F4B123CCA5D0C2F1BE4CA7B63ED7</Fingerprint>
     <Path>this-value-unused</Path>
    </PublicKey>
   </PublicKeys>
  </SSH>
  </LinuxProvisioningConfigurationSet>
</wa:ProvisioningSection>

hostname

When the user launches an instance, they provide a hostname for that instance. The hostname is provided to the instance in the ovf-env.xml file as HostName.

Whatever value the instance provides in its dhcp request will resolve in the domain returned in the ‘search’ request.

A generic image will already have a hostname configured. The ubuntu cloud images have ‘ubuntu’ as the hostname of the system, and the initial dhcp request on eth0 is not guaranteed to occur after the datasource code has been run. So, on first boot, that initial value will be sent in the dhcp request and that value will resolve.

In order to make the HostName provided in the ovf-env.xml resolve, a dhcp request must be made with the new value. cloud-init handles this by setting the hostname in the DataSource’s ‘get_data’ method via ‘hostname $HostName’, and then bouncing the interface. This behavior can be configured or disabled in the datasource config. See ‘Configuration’ above.

CloudSigma

This datasource finds metadata and user-data from the CloudSigma cloud platform. Data transfer occurs through a virtual serial port of the CloudSigma’s VM and the presence of network adapter is NOT a requirement, See server context in the public documentation for more information.

Setting a hostname

By default the name of the server will be applied as a hostname on the first boot.

Providing user-data

You can provide user-data to the VM using the dedicated meta field in the server context cloudinit-user-data. By default cloud-config format is expected there and the #cloud-config header could be omitted. However since this is a raw-text field you could provide any of the valid config formats.

You have the option to encode your user-data using Base64. In order to do that you have to add the cloudinit-user-data field to the base64_fields. The latter is a comma-separated field with all the meta fields whit base64 encoded values.

If your user-data does not need an internet connection you can create a meta field in the server context cloudinit-dsmode and set “local” as value. If this field does not exist the default value is “net”.

CloudStack

Apache CloudStack expose user-data, meta-data, user password and account SSH key thru the Virtual-Router. The datasource obtains the VR address via dhcp lease information given to the instance. For more details on meta-data and user-data, refer the CloudStack Administrator Guide.

URLs to access user-data and meta-data from the Virtual Machine. data-server. is a well-known hostname provided by the CloudStack virtual router that points to the next UserData server (which is usually also the virtual router).

http://data-server./latest/user-data
http://data-server./latest/meta-data
http://data-server./latest/meta-data/{metadata type}

If data-server. cannot be resolved, cloud-init will try to obtain the virtual router’s address from the system’s DHCP leases. If that fails, it will use the system’s default gateway.

Configuration

The following configuration can be set for the datasource in system configuration (in /etc/cloud/cloud.cfg or /etc/cloud/cloud.cfg.d/).

The settings that may be configured are:

• max_wait: the maximum amount of clock time in seconds that should be spent searching metadata_urls. A value less than zero will result in only one request being made, to the first in the list. (default: 120)

• timeout: the timeout value provided to urlopen for each individual http request. This is used both when selecting a metadata_url and when crawling the metadata service. (default: 50)

An example configuration with the default values is provided below:

datasource:
  CloudStack:
    max_wait: 120
    timeout: 50

Config Drive

The configuration drive datasource supports the OpenStack configuration drive disk.

See the config drive extension and metadata introduction in the public documentation for more information.

By default, cloud-init does always consider this source to be a full-fledged datasource. Instead, the typical behavior is to assume it is really only present to provide networking information. Cloud-init will copy off the network information, apply it to the system, and then continue on. The “full” datasource could then be found in the EC2 metadata service. If this is not the case then the files contained on the located drive must provide equivalents to what the EC2 metadata service would provide (which is typical of the version 2 support listed below)

Version 1 (Deprecated)

Note: Version 1 is legacy and should be considered deprecated. Version 2 has been supported in OpenStack since 2012.2 (Folsom).

The following criteria are required to as a config drive:

1. Must be formatted with vfat filesystem

2. Must contain one of the following files

/etc/network/interfaces
/root/.ssh/authorized_keys
/meta.js

/etc/network/interfaces

This file is laid down by nova in order to pass static networking information to the guest. Cloud-init will copy it off of the config-drive and into /etc/network/interfaces (or convert it to RH format) as soon as it can, and then attempt to bring up all network interfaces.

/root/.ssh/authorized_keys

This file is laid down by nova, and contains the ssk keys that were provided to nova on instance creation (nova-boot –key ….)

/meta.js

meta.js is populated on the config-drive in response to the user passing “meta flags” (nova boot –meta key=value …). It is expected to be json formatted.

Version 2

The following criteria are required to as a config drive:

1. Must be formatted with vfat or iso9660 filesystem or have a filesystem label of config-2 or CONFIG-2

2. The files that will typically be present in the config drive are:

openstack/
  - 2012-08-10/ or latest/
    - meta_data.json
    - user_data (not mandatory)
  - content/
    - 0000 (referenced content files)
    - 0001
    - ....
ec2
  - latest/
    - meta-data.json (not mandatory)

Keys and values

Cloud-init’s behavior can be modified by keys found in the meta.js (version 1 only) file in the following ways.

dsmode:
  values: local, net, pass
  default: pass

This is what indicates if configdrive is a final data source or not. By default it is ‘pass’, meaning this datasource should not be read. Set it to ‘local’ or ‘net’ to stop cloud-init from continuing on to search for other data sources after network config.

The difference between ‘local’ and ‘net’ is that local will not require networking to be up before user-data actions (or boothooks) are run.

instance-id:
  default: iid-dsconfigdrive

This is utilized as the metadata’s instance-id. It should generally be unique, as it is what is used to determine “is this a new instance”.

public-keys:
  default: None

If present, these keys will be used as the public keys for the instance. This value overrides the content in authorized_keys.

Note: it is likely preferable to provide keys via user-data

user-data:
  default: None

This provides cloud-init user-data. See examples for what all can be present here.

DigitalOcean

The DigitalOcean datasource consumes the content served from DigitalOcean’s metadata service. This metadata service serves information about the running droplet via HTTP over the link local address 169.254.169.254. The metadata API endpoints are fully described at https://developers.digitalocean.com/metadata/.

Configuration

DigitalOcean’s datasource can be configured as follows:

datasource:

DigitalOcean:

retries: 3 timeout: 2

• retries: Determines the number of times to attempt to connect to the metadata service

• timeout: Determines the timeout in seconds to wait for a response from the metadata service

E24Cloud

E24Cloud platform provides an AWS Ec2 metadata service clone. It identifies itself to guests using the dmi system-manufacturer (/sys/class/dmi/id/sys_vendor).

Exoscale

This datasource supports reading from the metadata server used on the Exoscale platform.

Use of the Exoscale datasource is recommended to benefit from new features of the Exoscale platform.

The datasource relies on the availability of a compatible metadata server (http://169.254.169.254 is used by default) and its companion password server, reachable at the same address (by default on port 8080).

Crawling of metadata

The metadata service and password server are crawled slightly differently:

• The “metadata service” is crawled every boot.

• The password server is also crawled every boot (the Exoscale datasource forces the password module to run with “frequency always”).

In the password server case, the following rules apply in order to enable the “restore instance password” functionality:

• If a password is returned by the password server, it is then marked “saved” by the cloud-init datasource. Subsequent boots will skip setting the password (the password server will return “saved_password”).

• When the instance password is reset (via the Exoscale UI), the password server will return the non-empty password at next boot, therefore causing cloud-init to reset the instance’s password.

Configuration

Users of this datasource are discouraged from changing the default settings unless instructed to by Exoscale support.

The following settings are available and can be set for the datasource in system configuration (in /etc/cloud/cloud.cfg.d/).

The settings available are:

• metadata_url: The URL for the metadata service (defaults to http://169.254.169.254)

• api_version: The API version path on which to query the instance metadata (defaults to 1.0)

• password_server_port: The port (on the metadata server) on which the password server listens (defaults to 8080).

• timeout: the timeout value provided to urlopen for each individual http request. (defaults to 10)

• retries: The number of retries that should be done for an http request (defaults to 6)

An example configuration with the default values is provided below:

datasource:
  Exoscale:
    metadata_url: "http://169.254.169.254"
    api_version: "1.0"
    password_server_port: 8080
    timeout: 10
    retries: 6

Fallback/None

This is the fallback datasource when no other datasource can be selected. It is the equivalent of a empty datasource in that it provides a empty string as userdata and a empty dictionary as metadata. It is useful for testing as well as for when you do not have a need to have an actual datasource to meet your instance requirements (ie you just want to run modules that are not concerned with any external data). It is typically put at the end of the datasource search list so that if all other datasources are not matched, then this one will be so that the user is not left with an inaccessible instance.

Note: the instance id that this datasource provides is iid-datasource-none.

Google Compute Engine

The GCE datasource gets its data from the internal compute metadata server. Metadata can be queried at the URL ‘http://metadata.google.internal/computeMetadata/v1/’ from within an instance. For more information see the GCE metadata docs.

Currently the default project and instance level metadatakeys keys project/attributes/sshKeys and instance/attributes/ssh-keys are merged to provide public-keys.

user-data and user-data-encoding can be provided to cloud-init by setting those custom metadata keys for an instance.

Configuration

The following configuration can be set for the datasource in system configuration (in /etc/cloud/cloud.cfg or /etc/cloud/cloud.cfg.d/).

The settings that may be configured are:

• retries: The number of retries that should be done for an http request. This value is used only after metadata_url is selected. (default: 5)

• sec_between_retries: The amount of wait time between the retries when crawling the metadata service. (default: 1)

An example configuration with the default values is provided below:

datasource:
  GCE:
    retries: 5
    sec_between_retries: 1

LXD

The data source LXD allows the user to provide custom user-data, vendor-data, meta-data and network-config to the instance without running a network service (or even without having a network at all). This datasource performs HTTP GETs against the LXD socket device which is provided to each running LXD container and VM as /dev/lxd/sock and represents all instance-metadata as versioned HTTP routes such as:

• 1.0/meta-data

• 1.0/config/user.meta-data

• 1.0/config/user.vendor-data

• 1.0/config/user.user-data

• 1.0/config/user.<any-custom-key>

The LXD socket device /dev/lxd/sock is only present on containers and VMs when the instance configuration has security.devlxd=true (default). Disabling security.devlxd configuration setting at initial launch will ensure that cloud-init uses the NoCloud datasource. Disabling security.devlxd ove the life of the container will result in warnings from cloud-init and cloud-init will keep the originally detected LXD datasource.

The LXD datasource provides cloud-init the opportunity to react to meta-data, vendor-data, user-data and network-config changes and render the updated configuration across a system reboot.

One can manipulate what meta-data, vendor-data or user-data is provided to the launched container using the LXD profiles or lxc launch ... -c <key>="<value>" at initial container launch using one of the following keys:

• user.meta-data: YAML metadata which will be appended to base meta-data

• user.vendor-data: YAML which overrides any meta-data values

• user.network-config: YAML representing either Networking Config Version 1 or Networking Config Version 2 format

• user.user-data: YAML which takes preference and overrides both meta-data and vendor-data values

• user.any-key: Custom user configuration key and value pairs can be passed to cloud-init. Those keys/values will be present in instance-data which can be used by both #template: jinja #cloud-config templates and the cloud-init query command.

By default, network configuration from this datasource will be:

version: 1
config:
    - type: physical
      name: eth0
      subnets:
          - type: dhcp
            control: auto

This datasource is intended to replace NoCloud datasource for LXD instances with a more direct support for LXD APIs instead of static NoCloud seed files.

MAAS

For now see: https://maas.io/docs

NoCloud

The data source NoCloud allows the user to provide user-data and meta-data to the instance without running a network service (or even without having a network at all).

You can provide meta-data and user-data to a local vm boot via files on a vfat or iso9660 filesystem. The filesystem volume label must be cidata or CIDATA.

Alternatively, you can provide meta-data via kernel command line or SMBIOS “serial number” option. The data must be passed in the form of a string:

ds=nocloud[;key=val;key=val]

or

ds=nocloud-net[;key=val;key=val]

The permitted keys are:

• h or local-hostname

• i or instance-id

• s or seedfrom

With ds=nocloud, the seedfrom value must start with / or file://. With ds=nocloud-net, the seedfrom value must start with http:// or https://.

e.g. you can pass this option to QEMU:

-smbios type=1,serial=ds=nocloud-net;s=http://10.10.0.1:8000/

to cause NoCloud to fetch the full meta-data from http://10.10.0.1:8000/meta-data after the network initialization is complete.

These user-data and meta-data files are expected to be in the following format.

/user-data
/meta-data

Both files are required to be present for it to be considered a valid seed ISO.

Basically, user-data is simply user-data and meta-data is a YAML formatted file representing what you’d find in the EC2 metadata service.

You may also optionally provide a vendor-data file in the following format.

/vendor-data

Given a disk ubuntu 12.04 cloud image in ‘disk.img’, you can create a sufficient disk by following the example below.

## create user-data and meta-data files that will be used
## to modify image on first boot
$ { echo instance-id: iid-local01; echo local-hostname: cloudimg; } > meta-data

$ printf "#cloud-config\npassword: passw0rd\nchpasswd: { expire: False }\nssh_pwauth: True\n" > user-data

## create a disk to attach with some user-data and meta-data
$ genisoimage  -output seed.iso -volid cidata -joliet -rock user-data meta-data

## alternatively, create a vfat filesystem with same files
## $ truncate --size 2M seed.img
## $ mkfs.vfat -n cidata seed.img
## $ mcopy -oi seed.img user-data meta-data ::

## create a new qcow image to boot, backed by your original image
$ qemu-img create -f qcow2 -b disk.img boot-disk.img

## boot the image and login as 'ubuntu' with password 'passw0rd'
## note, passw0rd was set as password through the user-data above,
## there is no password set on these images.
$ kvm -m 256 \
   -net nic -net user,hostfwd=tcp::2222-:22 \
   -drive file=boot-disk.img,if=virtio \
   -drive file=seed.iso,if=virtio

Note: that the instance-id provided (iid-local01 above) is what is used to determine if this is “first boot”. So if you are making updates to user-data you will also have to change that, or start the disk fresh.

Also, you can inject an /etc/network/interfaces file by providing the content for that file in the network-interfaces field of metadata.

Example metadata:

instance-id: iid-abcdefg
network-interfaces: |
  iface eth0 inet static
  address 192.168.1.10
  network 192.168.1.0
  netmask 255.255.255.0
  broadcast 192.168.1.255
  gateway 192.168.1.254
hostname: myhost

Network configuration can also be provided to cloud-init in either Networking Config Version 1 or Networking Config Version 2 by providing that YAML formatted data in a file named network-config. If found, this file will override a network-interfaces file.

See an example below. Note specifically that this file does not have a top level network key as it is already assumed to be network configuration based on the filename.

version: 1
config:
   - type: physical
     name: interface0
     mac_address: "52:54:00:12:34:00"
     subnets:
        - type: static
          address: 192.168.1.10
          netmask: 255.255.255.0
          gateway: 192.168.1.254

version: 2
ethernets:
  interface0:
    match:
      mac_address: "52:54:00:12:34:00"
    set-name: interface0
    addresses:
      - 192.168.1.10/255.255.255.0
    gateway4: 192.168.1.254

OpenNebula

The OpenNebula (ON) datasource supports the contextualization disk.

See contextualization overview, contextualizing VMs and network configuration in the public documentation for more information.

OpenNebula’s virtual machines are contextualized (parametrized) by CD-ROM image, which contains a shell script context.sh with custom variables defined on virtual machine start. There are no fixed contextualization variables, but the datasource accepts many used and recommended across the documentation.

Datasource configuration

Datasource accepts following configuration options.

dsmode:
  values: local, net, disabled
  default: net

Tells if this datasource will be processed in ‘local’ (pre-networking) or ‘net’ (post-networking) stage or even completely ‘disabled’.

parseuser:
  default: nobody

Unprivileged system user used for contextualization script processing.

Contextualization disk

The following criteria are required:

1. Must be formatted with iso9660 filesystem or have a filesystem label of CONTEXT or CDROM

2. Must contain file context.sh with contextualization variables. File is generated by OpenNebula, it has a KEY=’VALUE’ format and can be easily read by bash

Contextualization variables

There are no fixed contextualization variables in OpenNebula, no standard. Following variables were found on various places and revisions of the OpenNebula documentation. Where multiple similar variables are specified, only first found is taken.

DSMODE

Datasource mode configuration override. Values: local, net, disabled.

DNS
ETH<x>_IP
ETH<x>_NETWORK
ETH<x>_MASK
ETH<x>_GATEWAY
ETH<x>_GATEWAY6
ETH<x>_DOMAIN
ETH<x>_DNS
ETH<x>_SEARCH_DOMAIN
ETH<x>_MTU
ETH<x>_IP6
ETH<x>_IP6_ULA
ETH<x>_IP6_PREFIX_LENGTH
ETH<x>_IP6_GATEWAY

Static network configuration.

SET_HOSTNAME
HOSTNAME

Instance hostname.

PUBLIC_IP
IP_PUBLIC
ETH0_IP

If no hostname has been specified, cloud-init will try to create hostname from instance’s IP address in ‘local’ dsmode. In ‘net’ dsmode, cloud-init tries to resolve one of its IP addresses to get hostname.

SSH_KEY
SSH_PUBLIC_KEY

One or multiple SSH keys (separated by newlines) can be specified.

USER_DATA
USERDATA

cloud-init user data.

Example configuration

This example cloud-init configuration (cloud.cfg) enables OpenNebula datasource only in ‘net’ mode.

disable_ec2_metadata: True
datasource_list: ['OpenNebula']
datasource:
  OpenNebula:
    dsmode: net
    parseuser: nobody

Example VM’s context section

CONTEXT=[
  SSH_KEY="$USER[SSH_KEY]
$USER[SSH_KEY1]
$USER[SSH_KEY2]",
  PUBLIC_IP="$NIC[IP]",
  USER_DATA="#cloud-config
# see https://help.ubuntu.com/community/CloudInit

packages: []

mounts:
- [vdc,none,swap,sw,0,0]
runcmd:
- echo 'Instance has been configured by cloud-init.' | wall
" ]

OpenStack

This datasource supports reading data from the OpenStack Metadata Service.

Discovery

To determine whether a platform looks like it may be OpenStack, cloud-init checks the following environment attributes as a potential OpenStack platform:

• Maybe OpenStack if:

  • non-x86 cpu architecture: because DMI data is buggy on some arches

• Is OpenStack if x86 architecture and ANY of the following:

  • /proc/1/environ: Nova-lxd contains product_name=OpenStack Nova

  • DMI product_name: Either Openstack Nova or OpenStack Compute

  • DMI chassis_asset_tag is OpenTelekomCloud, SAP CCloud VM, OpenStack Nova (since 19.2) or OpenStack Compute (since 19.2)

Configuration

The following configuration can be set for the datasource in system configuration (in /etc/cloud/cloud.cfg or /etc/cloud/cloud.cfg.d/).

The settings that may be configured are:

• metadata_urls: This list of urls will be searched for an OpenStack metadata service. The first entry that successfully returns a 200 response for <url>/openstack will be selected. (default: [‘http://169.254.169.254’])

• max_wait: the maximum amount of clock time in seconds that should be spent searching metadata_urls. A value less than zero will result in only one request being made, to the first in the list. (default: -1)

• timeout: the timeout value provided to urlopen for each individual http request. This is used both when selecting a metadata_url and when crawling the metadata service. (default: 10)

• retries: The number of retries that should be done for an http request. This value is used only after metadata_url is selected. (default: 5)

• apply_network_config: A boolean specifying whether to configure the network for the instance based on network_data.json provided by the metadata service. When False, only configure dhcp on the primary nic for this instances. (default: True)

An example configuration with the default values is provided below:

datasource:
  OpenStack:
    metadata_urls: ["http://169.254.169.254"]
    max_wait: -1
    timeout: 10
    retries: 5
    apply_network_config: True

Vendor Data

The OpenStack metadata server can be configured to serve up vendor data which is available to all instances for consumption. OpenStack vendor data is, generally, a JSON object.

cloud-init will look for configuration in the cloud-init attribute of the vendor data JSON object. cloud-init processes this configuration using the same handlers as user data, so any formats that work for user data should work for vendor data.

For example, configuring the following as vendor data in OpenStack would upgrade packages and install htop on all instances:

{"cloud-init": "#cloud-config\npackage_upgrade: True\npackages:\n - htop"}

For more general information about how cloud-init handles vendor data, including how it can be disabled by users on instances, see Vendor Data.

OpenStack can also be configured to provide ‘dynamic vendordata’ which is provided by the DynamicJSON provider and appears under a different metadata path, /vendor_data2.json.

Cloud-init will look for a cloud-init at the vendor_data2 path; if found, settings are applied after (and, hence, overriding) the settings from static vendor data. Both sets of vendor data can be overridden by user data.

Oracle

This datasource reads metadata, vendor-data and user-data from Oracle Compute Infrastructure (OCI).

Oracle Platform

OCI provides bare metal and virtual machines. In both cases, the platform identifies itself via DMI data in the chassis asset tag with the string ‘OracleCloud.com’.

Oracle’s platform provides a metadata service that mimics the 2013-10-17 version of OpenStack metadata service. Initially support for Oracle was done via the OpenStack datasource.

Cloud-init has a specific datasource for Oracle in order to:

1. allow and support future growth of the OCI platform.

2. address small differences between OpenStack and Oracle metadata implementation.

Configuration

The following configuration can be set for the datasource in system configuration (in /etc/cloud/cloud.cfg or /etc/cloud/cloud.cfg.d/).

The settings that may be configured are:

• configure_secondary_nics: A boolean, defaulting to False. If set to True on an OCI Virtual Machine, cloud-init will fetch networking metadata from Oracle’s IMDS and use it to configure the non-primary network interface controllers in the system. If set to True on an OCI Bare Metal Machine, it will have no effect (though this may change in the future).

An example configuration with the default values is provided below:

datasource:
 Oracle:
  configure_secondary_nics: false

OVF

The OVF Datasource provides a datasource for reading data from on an Open Virtualization Format ISO transport.

For further information see a full working example in cloud-init’s source code tree in doc/sources/ovf

Configuration

The following configuration can be set for the datasource in system configuration (in /etc/cloud/cloud.cfg or /etc/cloud/cloud.cfg.d/).

The settings that may be configured are:

• disable_vmware_customization: disable or enable the vmware customization based on vmware customization files. (default: True)

• allow_raw_data: enable or disable the vmware customization based on raw cloud-init data including metadata and userdata. (default: True)

• vmware_cust_file_max_wait: the maximum amount of clock time in seconds that should be spent waiting for vmware customization files. (default: 15)

On VMware platforms, VMTools use is required for OVF datasource configuration settings as well as vCloud and vSphere admin configuration. User could change the VMTools configuration options with command:

vmware-toolbox-cmd config set <section> <key> <value>

The following VMTools configuration options affect cloud-init’s behavior on a booted VM:

• a: [deploypkg] enable-custom-scripts

  If this option is absent in VMTools configuration, the custom script is disabled by default for security reasons. Some VMware products could change this default behavior (for example: enabled by default) via customization specification settings.

VMWare admin can refer to (https://github.com/canonical/cloud-init/blob/main/cloudinit/sources/helpers/vmware/imc/config.py) and set the customization specification settings.

For more information, see VMware vSphere Product Documentation and specific VMTools parameters consumed.

Rbx Cloud

The Rbx datasource consumes the metadata drive available on platform HyperOne and Rootbox platform.

Datasource supports, in particular, network configurations, hostname, user accounts and user metadata.

Metadata drive

Drive metadata is a FAT-formatted partition with the `CLOUDMD` or `cloudmd` label on the system disk. Its contents are refreshed each time the virtual machine is restarted, if the partition exists. For more information see HyperOne Virtual Machine docs.

SmartOS Datasource

This datasource finds metadata and user-data from the SmartOS virtualization platform (i.e. Joyent).

Please see http://smartos.org/ for information about SmartOS.

SmartOS Platform

The SmartOS virtualization platform uses meta-data to the instance via the second serial console. On Linux, this is /dev/ttyS1. The data is a provided via a simple protocol: something queries for the data, the console responds with the status and if “SUCCESS” returns until a single “.n”.

New versions of the SmartOS tooling will include support for base64 encoded data.

Meta-data channels

Cloud-init supports three modes of delivering user/meta-data via the flexible channels of SmartOS.

• user-data is written to /var/db/user-data

  • per the spec, user-data is for consumption by the end-user, not provisioning tools

  • cloud-init entirely ignores this channel other than writing it to disk

  • removal of the meta-data key means that /var/db/user-data gets removed

  • a backup of previous meta-data is maintained as /var/db/user-data.<timestamp>. <timestamp> is the epoch time when cloud-init ran

• user-script is written to /var/lib/cloud/scripts/per-boot/99_user_data

  • this is executed each boot

  • a link is created to /var/db/user-script

  • previous versions of the user-script is written to /var/lib/cloud/scripts/per-boot.backup/99_user_script.<timestamp>. - <timestamp> is the epoch time when cloud-init ran.

  • when the ‘user-script’ meta-data key goes missing, the user-script is removed from the file system, although a backup is maintained.

  • if the script does not start with a shebang (i.e. starts with #!<executable>), then or is not an executable, cloud-init will add a shebang of “#!/bin/bash”

• cloud-init:user-data is treated like on other Clouds.

  • this channel is used for delivering _all_ cloud-init instructions

  • scripts delivered over this channel must be well formed (i.e. must have a shebang)

Cloud-init supports reading the traditional meta-data fields supported by the SmartOS tools. These are:

• root_authorized_keys

• hostname

• enable_motd_sys_info

• iptables_disable

Note: At this time iptables_disable and enable_motd_sys_info are read but

are not actioned.

Disabling user-script

Cloud-init uses the per-boot script functionality to handle the execution of the user-script. If you want to prevent this use a cloud-config of:

#cloud-config
cloud_final_modules:
 - scripts-per-once
 - scripts-per-instance
 - scripts-user
 - ssh-authkey-fingerprints
 - keys-to-console
 - phone-home
 - final-message
 - power-state-change

Alternatively you can use the json patch method

#cloud-config-jsonp
[
     { "op": "replace",
       "path": "/cloud_final_modules",
       "value": ["scripts-per-once",
                 "scripts-per-instance",
                 "scripts-user",
                 "ssh-authkey-fingerprints",
                 "keys-to-console",
                 "phone-home",
                 "final-message",
                 "power-state-change"]
     }
]

The default cloud-config includes “script-per-boot”. Cloud-init will still ingest and write the user-data but will not execute it, when you disable the per-boot script handling.

Note: Unless you have an explicit use-case, it is recommended that you not

disable the per-boot script execution, especially if you are using any of the life-cycle management features of SmartOS.

The cloud-config needs to be delivered over the cloud-init:user-data channel in order for cloud-init to ingest it.

base64

The following are exempt from base64 encoding, owing to the fact that they are provided by SmartOS:

• root_authorized_keys

• enable_motd_sys_info

• iptables_disable

• user-data

• user-script

This list can be changed through system config of variable ‘no_base64_decode’.

This means that user-script and user-data as well as other values can be base64 encoded. Since Cloud-init can only guess as to whether or not something is truly base64 encoded, the following meta-data keys are hints as to whether or not to base64 decode something:

• base64_all: Except for excluded keys, attempt to base64 decode the values. If the value fails to decode properly, it will be returned in its text

• base64_keys: A comma delimited list of which keys are base64 encoded.

• b64-<key>: for any key, if there exists an entry in the metadata for ‘b64-<key>’ Then ‘b64-<key>’ is expected to be a plaintext boolean indicating whether or not its value is encoded.

• no_base64_decode: This is a configuration setting (i.e. /etc/cloud/cloud.cfg.d) that sets which values should not be base64 decoded.

disk_aliases and ephemeral disk

By default, SmartOS only supports a single ephemeral disk. That disk is completely empty (un-partitioned with no filesystem).

The SmartOS datasource has built-in cloud-config which instructs the ‘disk_setup’ module to partition and format the ephemeral disk.

You can control the disk_setup then in 2 ways:

1. through the datasource config, you can change the ‘alias’ of ephermeral0 to reference another device. The default is:

   ‘disk_aliases’: {‘ephemeral0’: ‘/dev/vdb’},

   Which means anywhere disk_setup sees a device named ‘ephemeral0’ then /dev/vdb will be substituted.

2. you can provide disk_setup or fs_setup data in user-data to overwrite the datasource’s built-in values.

See doc/examples/cloud-config-disk-setup.txt for information on disk_setup.

UpCloud

The UpCloud datasource consumes information from UpCloud’s metadata service. This metadata service serves information about the running server via HTTP over the address 169.254.169.254 available in every DHCP-configured interface. The metadata API endpoints are fully described in UpCloud API documentation at https://developers.upcloud.com/1.3/8-servers/#metadata-service.

Providing user-data

When creating a server, user-data is provided by specifying it as user_data in the API or via the server creation tool in the control panel. User-data is immutable during server’s lifetime and can be removed by deleting the server.

VMware

This datasource is for use with systems running on a VMware platform such as vSphere and currently supports the following data transports:

• GuestInfo keys

Configuration

The configuration method is dependent upon the transport:

GuestInfo Keys

One method of providing meta, user, and vendor data is by setting the following key/value pairs on a VM’s extraConfig property:

Property	Description
guestinfo.metadata	A YAML or JSON document containing the cloud-init metadata.
guestinfo.metadata.encoding	The encoding type for guestinfo.metadata.
guestinfo.userdata	A YAML document containing the cloud-init user data.
guestinfo.userdata.encoding	The encoding type for guestinfo.userdata.
guestinfo.vendordata	A YAML document containing the cloud-init vendor data.
guestinfo.vendordata.encoding	The encoding type for guestinfo.vendordata.

All guestinfo.*.encoding values may be set to base64 or gzip+base64.

Features

This section reviews several features available in this datasource, regardless of how the meta, user, and vendor data was discovered.

Instance data and lazy networks

One of the hallmarks of cloud-init is its use of instance-data and JINJA queries – the ability to write queries in user and vendor data that reference runtime information present in /run/cloud-init/instance-data.json. This works well when the metadata provides all of the information up front, such as the network configuration. For systems that rely on DHCP, however, this information may not be available when the metadata is persisted to disk.

This datasource ensures that even if the instance is using DHCP to configure networking, the same details about the configured network are available in /run/cloud-init/instance-data.json as if static networking was used. This information collected at runtime is easy to demonstrate by executing the datasource on the command line. From the root of this repository, run the following command:

PYTHONPATH="$(pwd)" python3 cloudinit/sources/DataSourceVMware.py

The above command will result in output similar to the below JSON:

{
    "hostname": "akutz.localhost",
    "local-hostname": "akutz.localhost",
    "local-ipv4": "192.168.0.188",
    "local_hostname": "akutz.localhost",
    "network": {
        "config": {
            "dhcp": true
        },
        "interfaces": {
            "by-ipv4": {
                "172.0.0.2": {
                    "netmask": "255.255.255.255",
                    "peer": "172.0.0.2"
                },
                "192.168.0.188": {
                    "broadcast": "192.168.0.255",
                    "mac": "64:4b:f0:18:9a:21",
                    "netmask": "255.255.255.0"
                }
            },
            "by-ipv6": {
                "fd8e:d25e:c5b6:1:1f5:b2fd:8973:22f2": {
                    "flags": 208,
                    "mac": "64:4b:f0:18:9a:21",
                    "netmask": "ffff:ffff:ffff:ffff::/64"
                }
            },
            "by-mac": {
                "64:4b:f0:18:9a:21": {
                    "ipv4": [
                        {
                            "addr": "192.168.0.188",
                            "broadcast": "192.168.0.255",
                            "netmask": "255.255.255.0"
                        }
                    ],
                    "ipv6": [
                        {
                            "addr": "fd8e:d25e:c5b6:1:1f5:b2fd:8973:22f2",
                            "flags": 208,
                            "netmask": "ffff:ffff:ffff:ffff::/64"
                        }
                    ]
                },
                "ac:de:48:00:11:22": {
                    "ipv6": []
                }
            }
        }
    },
    "wait-on-network": {
        "ipv4": true,
        "ipv6": "false"
    }
}

Redacting sensitive information

Sometimes the cloud-init userdata might contain sensitive information, and it may be desirable to have the guestinfo.userdata key (or other guestinfo keys) redacted as soon as its data is read by the datasource. This is possible by adding the following to the metadata:

redact: # formerly named cleanup-guestinfo, which will also work
- userdata
- vendordata

When the above snippet is added to the metadata, the datasource will iterate over the elements in the redact array and clear each of the keys. For example, when the guestinfo transport is used, the above snippet will cause the following commands to be executed:

vmware-rpctool "info-set guestinfo.userdata ---"
vmware-rpctool "info-set guestinfo.userdata.encoding  "
vmware-rpctool "info-set guestinfo.vendordata ---"
vmware-rpctool "info-set guestinfo.vendordata.encoding  "

Please note that keys are set to the valid YAML string --- as it is not possible remove an existing key from the guestinfo key-space. A key’s analogous encoding property will be set to a single white-space character, causing the datasource to treat the actual key value as plain-text, thereby loading it as an empty YAML doc (hence the aforementioned ---).

Reading the local IP addresses

This datasource automatically discovers the local IPv4 and IPv6 addresses for a guest operating system based on the default routes. However, when inspecting a VM externally, it’s not possible to know what the default IP address is for the guest OS. That’s why this datasource sets the discovered, local IPv4 and IPv6 addresses back in the guestinfo namespace as the following keys:

• guestinfo.local-ipv4

• guestinfo.local-ipv6

It is possible that a host may not have any default, local IP addresses. It’s also possible the reported, local addresses are link-local addresses. But these two keys may be used to discover what this datasource determined were the local IPv4 and IPv6 addresses for a host.

Waiting on the network

Sometimes cloud-init may bring up the network, but it will not finish coming online before the datasource’s setup function is called, resulting in an /var/run/cloud-init/instance-data.json file that does not have the correct network information. It is possible to instruct the datasource to wait until an IPv4 or IPv6 address is available before writing the instance data with the following metadata properties:

wait-on-network:
  ipv4: true
  ipv6: true

If either of the above values are true, then the datasource will sleep for a second, check the network status, and repeat until one or both addresses from the specified families are available.

Walkthrough

The following series of steps is a demonstration on how to configure a VM with this datasource:

1. Create the metadata file for the VM. Save the following YAML to a file named metadata.yaml:

   instance-id: cloud-vm
   local-hostname: cloud-vm
   network:
     version: 2
     ethernets:
       nics:
         match:
           name: ens*
         dhcp4: yes
   

2. Create the userdata file userdata.yaml:

   #cloud-config
   
   users:
   - default
   - name: akutz
       primary_group: akutz
       sudo: ALL=(ALL) NOPASSWD:ALL
       groups: sudo, wheel
       lock_passwd: true
       ssh_authorized_keys:
       - ssh-rsa AAAAB3NzaC1yc2EAAAADAQABAAACAQDE0c5FczvcGSh/tG4iw+Fhfi/O5/EvUM/96js65tly4++YTXK1d9jcznPS5ruDlbIZ30oveCBd3kT8LLVFwzh6hepYTf0YmCTpF4eDunyqmpCXDvVscQYRXyasEm5olGmVe05RrCJSeSShAeptv4ueIn40kZKOghinGWLDSZG4+FFfgrmcMCpx5YSCtX2gvnEYZJr0czt4rxOZuuP7PkJKgC/mt2PcPjooeX00vAj81jjU2f3XKrjjz2u2+KIt9eba+vOQ6HiC8c2IzRkUAJ5i1atLy8RIbejo23+0P4N2jjk17QySFOVHwPBDTYb0/0M/4ideeU74EN/CgVsvO6JrLsPBR4dojkV5qNbMNxIVv5cUwIy2ThlLgqpNCeFIDLCWNZEFKlEuNeSQ2mPtIO7ETxEL2Cz5y/7AIuildzYMc6wi2bofRC8HmQ7rMXRWdwLKWsR0L7SKjHblIwarxOGqLnUI+k2E71YoP7SZSlxaKi17pqkr0OMCF+kKqvcvHAQuwGqyumTEWOlH6TCx1dSPrW+pVCZSHSJtSTfDW2uzL6y8k10MT06+pVunSrWo5LHAXcS91htHV1M1UrH/tZKSpjYtjMb5+RonfhaFRNzvj7cCE1f3Kp8UVqAdcGBTtReoE8eRUT63qIxjw03a7VwAyB2w+9cu1R9/vAo8SBeRqw== sakutz@gmail.com
   

3. Please note this step requires that the VM be powered off. All of the commands below use the VMware CLI tool, govc.

   Go ahead and assign the path to the VM to the environment variable VM:

   export VM="/inventory/path/to/the/vm"
   

4. Power off the VM:

   ---

   ⚠️ First Boot Mode

   To ensure the next power-on operation results in a first-boot scenario for cloud-init, it may be necessary to run the following command just before powering off the VM:

   cloud-init clean
   

   Otherwise cloud-init may not run in first-boot mode. For more information on how the boot mode is determined, please see the First Boot Documentation.

   ---

   govc vm.power -off "${VM}"
   

5. Export the environment variables that contain the cloud-init metadata and userdata:

   export METADATA=$(gzip -c9 <metadata.yaml | { base64 -w0 2>/dev/null || base64; }) \
        USERDATA=$(gzip -c9 <userdata.yaml | { base64 -w0 2>/dev/null || base64; })
   

6. Assign the metadata and userdata to the VM:

   govc vm.change -vm "${VM}" \
   -e guestinfo.metadata="${METADATA}" \
   -e guestinfo.metadata.encoding="gzip+base64" \
   -e guestinfo.userdata="${USERDATA}" \
   -e guestinfo.userdata.encoding="gzip+base64"
   

   Please note the above commands include specifying the encoding for the properties. This is important as it informs the datasource how to decode the data for cloud-init. Valid values for metadata.encoding and userdata.encoding include:

   • base64

   • gzip+base64

7. Power on the VM:

   govc vm.power -vm "${VM}" -on
   

If all went according to plan, the CentOS box is:

• Locked down, allowing SSH access only for the user in the userdata

• Configured for a dynamic IP address via DHCP

• Has a hostname of cloud-vm

Examples

This section reviews common configurations:

Setting the hostname

The hostname is set by way of the metadata key local-hostname.

Setting the instance ID

The instance ID may be set by way of the metadata key instance-id. However, if this value is absent then the instance ID is read from the file /sys/class/dmi/id/product_uuid.

Providing public SSH keys

The public SSH keys may be set by way of the metadata key public-keys-data. Each newline-terminated string will be interpreted as a separate SSH public key, which will be placed in distro’s default user’s ~/.ssh/authorized_keys. If the value is empty or absent, then nothing will be written to ~/.ssh/authorized_keys.

Configuring the network

The network is configured by setting the metadata key network with a value consistent with Network Config Versions 1 or 2, depending on the Linux distro’s version of cloud-init.

The metadata key network.encoding may be used to indicate the format of the metadata key “network”. Valid encodings are base64 and gzip+base64.

Vultr

The Vultr datasource retrieves basic configuration values from the locally accessible metadata service. All data is served over HTTP from the address 169.254.169.254. The endpoints are documented in https://www.vultr.com/metadata/

Configuration

Vultr’s datasource can be configured as follows:

datasource:

Vultr:

url: ‘http://169.254.169.254’ retries: 3 timeout: 2 wait: 2

• url: The URL used to acquire the metadata configuration from

• retries: Determines the number of times to attempt to connect to the metadata service

• timeout: Determines the timeout in seconds to wait for a response from the metadata service

• wait: Determines the timeout in seconds to wait before retrying after accessible failure

ZStack

ZStack platform provides a AWS Ec2 metadata service, but with different datasource identity. More information about ZStack can be found at ZStack.

Discovery

To determine whether a vm running on ZStack platform, cloud-init checks DMI information by ‘dmidecode -s chassis-asset-tag’, if the output ends with ‘.zstack.io’, it’s running on ZStack platform:

Metadata

Same as EC2, instance metadata can be queried at

GET http://169.254.169.254/2009-04-04/meta-data/
instance-id
local-hostname

Userdata

Same as EC2, instance userdata can be queried at

GET http://169.254.169.254/2009-04-04/user-data/
meta_data.json
user_data
password

Creation

The datasource objects have a few touch points with cloud-init. If you are interested in adding a new datasource for your cloud platform you will need to take care of the following items:

• Identify a mechanism for positive identification of the platform: It is good practice for a cloud platform to positively identify itself to the guest. This allows the guest to make educated decisions based on the platform on which it is running. On the x86 and arm64 architectures, many clouds identify themselves through DMI data. For example, Oracle’s public cloud provides the string ‘OracleCloud.com’ in the DMI chassis-asset field.

  cloud-init enabled images produce a log file with details about the platform. Reading through this log in /run/cloud-init/ds-identify.log may provide the information needed to uniquely identify the platform. If the log is not present, you can generate it by running from source ./tools/ds-identify or the installed location /usr/lib/cloud-init/ds-identify.

  The mechanism used to identify the platform will be required for the ds-identify and datasource module sections below.

• Add datasource module ``cloudinit/sources/DataSource<CloudPlatform>.py``: It is suggested that you start by copying one of the simpler datasources such as DataSourceHetzner.

• Add tests for datasource module: Add a new file with some tests for the module to cloudinit/sources/test_<yourplatform>.py. For example see cloudinit/sources/tests/test_oracle.py

• Update ds-identify: In systemd systems, ds-identify is used to detect which datasource should be enabled or if cloud-init should run at all. You’ll need to make changes to tools/ds-identify.

• Add tests for ds-identify: Add relevant tests in a new class to tests/unittests/test_ds_identify.py. You can use TestOracle as an example.

• Add your datasource name to the builtin list of datasources: Add your datasource module name to the end of the datasource_list entry in cloudinit/settings.py.

• Add your cloud platform to apport collection prompts: Update the list of cloud platforms in cloudinit/apport.py. This list will be provided to the user who invokes ubuntu-bug cloud-init.

• Enable datasource by default in ubuntu packaging branches: Ubuntu packaging branches contain a template file debian/cloud-init.templates that ultimately sets the default datasource_list when installed via package. This file needs updating when the commit gets into a package.

• Add documentation for your datasource: You should add a new file in doc/datasources/<cloudplatform>.rst

Vendor Data

Overview

Vendordata is data provided by the entity that launches an instance (for example, the cloud provider). This data can be used to customize the image to fit into the particular environment it is being run in.

Vendordata follows the same rules as user-data, with the following caveats:

1. Users have ultimate control over vendordata. They can disable its execution or disable handling of specific parts of multipart input.

2. By default it only runs on first boot

3. Vendordata can be disabled by the user. If the use of vendordata is required for the instance to run, then vendordata should not be used.

4. user supplied cloud-config is merged over cloud-config from vendordata.

Users providing cloud-config data can use the ‘#cloud-config-jsonp’ method to more finely control their modifications to the vendor supplied cloud-config. For example, if both vendor and user have provided ‘runcmd’ then the default merge handler will cause the user’s runcmd to override the one provided by the vendor. To append to ‘runcmd’, the user could better provide multipart input with a cloud-config-jsonp part like:

#cloud-config-jsonp
[{ "op": "add", "path": "/runcmd", "value": ["my", "command", "here"]}]

Further, we strongly advise vendors to not ‘be evil’. By evil, we mean any action that could compromise a system. Since users trust you, please take care to make sure that any vendordata is safe, atomic, idempotent and does not put your users at risk.

Input Formats

cloud-init will download and cache to filesystem any vendor-data that it finds. Vendordata is handled exactly like user-data. That means that the vendor can supply multipart input and have those parts acted on in the same way as user-data.

The only differences are:

• vendor-data-defined scripts are stored in a different location than user-data-defined scripts (to avoid namespace collision)

• user can disable part handlers by cloud-config settings. For example, to disable handling of ‘part-handlers’ in vendor-data, the user could provide user-data like this:

  #cloud-config
  vendordata: {excluded: 'text/part-handler'}
  

Examples

There are examples in the examples subdirectory.

Additionally, the ‘tools’ directory contains ‘write-mime-multipart’, which can be used to easily generate mime-multi-part files from a list of input files. That data can then be given to an instance.

See ‘write-mime-multipart –help’ for usage.

Network Configuration

• Default Behavior

• Disabling Network Configuration

• Fallback Networking

• Network Configuration Sources

• Network Configuration Outputs

• Network Output Policy

• Network Configuration Tools

• Examples

Default Behavior

Cloud-init ‘s searches for network configuration in order of increasing precedence; each item overriding the previous.

Datasource

For example, OpenStack may provide network config in the MetaData Service.

System Config

A network: entry in /etc/cloud/cloud.cfg.d/* configuration files.

Kernel Command Line

ip= or network-config=<Base64 encoded YAML config string>

User-data cannot change an instance’s network configuration. In the absence of network configuration in any of the above sources , Cloud-init will write out a network configuration that will issue a DHCP request on a “first” network interface.

Note

The network-config value is expected to be a Base64 encoded YAML string in Networking Config Version 1 or Networking Config Version 2 format. Optionally it can be compressed with gzip prior to Base64 encoding.

Disabling Network Configuration

Users may disable Cloud-init ‘s network configuration capability and rely on other methods, such as embedded configuration or other customizations.

Cloud-init supports the following methods for disabling cloud-init.

Kernel Command Line

Cloud-init will check additionally check for the parameter network-config=disabled which will automatically disable any network configuration.

Example disabling kernel command line entry:

network-config=disabled

cloud config

In the combined cloud-init configuration dictionary, merged from /etc/cloud/cloud.cfg and /etc/cloud/cloud.cfg.d/*:

network:
  config: disabled

If Cloud-init ‘s networking config has not been disabled, and no other network information is found, then it will proceed to generate a fallback networking configuration.

Disabling Network Activation

Some datasources may not be initialized until after network has been brought up. In this case, cloud-init will attempt to bring up the interfaces specified by the datasource metadata.

This behavior can be disabled in the cloud-init configuration dictionary, merged from /etc/cloud/cloud.cfg and /etc/cloud/cloud.cfg.d/*:

disable_network_activation: true

Fallback Network Configuration

Cloud-init will attempt to determine which of any attached network devices is most likely to have a connection and then generate a network configuration to issue a DHCP request on that interface.

Cloud-init runs during early boot and does not expect composed network devices (such as Bridges) to be available. Cloud-init does not consider the following interface devices as likely ‘first’ network interfaces for fallback configuration; they are filtered out from being selected.

• loopback: name=lo

• Virtual Ethernet: name=veth*

• Software Bridges: type=bridge

• Software VLANs: type=vlan

Cloud-init will prefer network interfaces that indicate they are connected via the Linux carrier flag being set. If no interfaces are marked connected, then all unfiltered interfaces are potential connections.

Of the potential interfaces, Cloud-init will attempt to pick the “right” interface given the information it has available.

Finally after selecting the “right” interface, a configuration is generated and applied to the system.

Note

PhotonOS disables fallback networking configuration by default leaving network unrendered when no other network config is provided. If fallback config is still desired on PhotonOS, it can be enabled by providing disable_fallback_netcfg: false in /etc/cloud/cloud.cfg:sys_config settings.

Network Configuration Sources

Cloud-init accepts a number of different network configuration formats in support of different cloud substrates. The Datasource for these clouds in Cloud-init will detect and consume Datasource-specific network configuration formats for use when writing an instance’s network configuration.

The following Datasources optionally provide network configuration:

• Config Drive

  • OpenStack Metadata Service Network

  • Network Configuration ENI (Legacy)

• DigitalOcean

  • DigitalOcean JSON metadata

• NoCloud

  • Networking Config Version 1

  • Networking Config Version 2

  • Network Configuration ENI (Legacy)

• OpenNebula

  • Network Configuration ENI (Legacy)

• OpenStack

  • Network Configuration ENI (Legacy)

  • OpenStack Metadata Service Network

• SmartOS Datasource

  • SmartOS JSON Metadata

• UpCloud

  • UpCloud JSON metadata

• Vultr

  • Vultr JSON metadata

For more information on network configuration formats

Network Configuration ENI (Legacy)

Cloud-init supports reading and writing network config in the ENI format which is consumed by the ifupdown tool to parse and apply network configuration.

As an input format this is legacy. In cases where ENI format is available and another format is also available, it will prefer to use the other format. This can happen in either NoCloud or OpenStack datasources.

Please reference existing documentation for the /etc/network/interfaces(5) format.

Networking Config Version 1

This network configuration format lets users customize their instance’s networking interfaces by assigning subnet configuration, virtual device creation (bonds, bridges, vlans) routes and DNS configuration.

Required elements of a Network Config Version 1 are config and version.

Cloud-init will read this format from system config. For example the following could be present in /etc/cloud/cloud.cfg.d/custom-networking.cfg:

network:
  version: 1
  config:
  - type: physical
    name: eth0
    subnets:
      - type: dhcp

The NoCloud datasource can also provide cloud-init networking configuration in this Format.

Configuration Types

Within the network config portion, users include a list of configuration types. The current list of support type values are as follows:

• Physical (physical)

• Bond (bond)

• Bridge (bridge)

• VLAN (vlan)

• Nameserver (nameserver)

• Route (route)

Physical, Bond, Bridge and VLAN types may also include IP configuration under the key subnets.

• Subnet/IP (subnets)

Physical

The physical type configuration represents a “physical” network device, typically Ethernet-based. At least one of these entries is required for external network connectivity. Type physical requires only one key: name. A physical device may contain some or all of the following keys:

name: <desired device name>

A devices name must be less than 15 characters. Names exceeding the maximum will be truncated. This is a limitation of the Linux kernel network-device structure.

mac_address: <MAC Address>

The MAC Address is a device unique identifier that most Ethernet-based network devices possess. Specifying a MAC Address is optional. Letters must be lowercase.

Note

MAC addresses must be strings. As MAC addresses which consist of only the digits 0-9 (i.e. no hex a-f) can be interpreted as a base 60 integer per the YAML 1.1 spec it is best practice to quote all MAC addresses to ensure they are parsed as strings regardless of value.

Note

Cloud-init will handle the persistent mapping between a device’s name and the mac_address.

mtu: <MTU SizeBytes>

The MTU key represents a device’s Maximum Transmission Unit, the largest size packet or frame, specified in octets (eight-bit bytes), that can be sent in a packet- or frame-based network. Specifying mtu is optional.

Note

The possible supported values of a device’s MTU is not available at configuration time. It’s possible to specify a value too large or to small for a device and may be ignored by the device.

Physical Example:

network:
  version: 1
  config:
    # Simple network adapter
    - type: physical
      name: interface0
      mac_address: '00:11:22:33:44:55'
    # Second nic with Jumbo frames
    - type: physical
      name: jumbo0
      mac_address: aa:11:22:33:44:55
      mtu: 9000
    # 10G pair
    - type: physical
      name: gbe0
      mac_address: cd:11:22:33:44:00
    - type: physical
      name: gbe1
      mac_address: cd:11:22:33:44:02

Bond

A bond type will configure a Linux software Bond with one or more network devices. A bond type requires the following keys:

name: <desired device name>

A devices name must be less than 15 characters. Names exceeding the maximum will be truncated. This is a limitation of the Linux kernel network-device structure.

mac_address: <MAC Address>

When specifying MAC Address on a bond this value will be assigned to the bond device and may be different than the MAC address of any of the underlying bond interfaces. Specifying a MAC Address is optional. If mac_address is not present, then the bond will use one of the MAC Address values from one of the bond interfaces.

Note

MAC addresses must be strings. As MAC addresses which consist of only the digits 0-9 (i.e. no hex a-f) can be interpreted as a base 60 integer per the YAML 1.1 spec it is best practice to quote all MAC addresses to ensure they are parsed as strings regardless of value.

bond_interfaces: <List of network device names>

The bond_interfaces key accepts a list of network device name values from the configuration. This list may be empty.

mtu: <MTU SizeBytes>

The MTU key represents a device’s Maximum Transmission Unit, the largest size packet or frame, specified in octets (eight-bit bytes), that can be sent in a packet- or frame-based network. Specifying mtu is optional.

Note

The possible supported values of a device’s MTU is not available at configuration time. It’s possible to specify a value too large or to small for a device and may be ignored by the device.

params: <Dictionary of key: value bonding parameter pairs>

The params key in a bond holds a dictionary of bonding parameters. This dictionary may be empty. For more details on what the various bonding parameters mean please read the Linux Kernel Bonding.txt.

Valid params keys are:

• active_slave: Set bond attribute

• ad_actor_key: Set bond attribute

• ad_actor_sys_prio: Set bond attribute

• ad_actor_system: Set bond attribute

• ad_aggregator: Set bond attribute

• ad_num_ports: Set bond attribute

• ad_partner_key: Set bond attribute

• ad_partner_mac: Set bond attribute

• ad_select: Set bond attribute

• ad_user_port_key: Set bond attribute

• all_slaves_active: Set bond attribute

• arp_all_targets: Set bond attribute

• arp_interval: Set bond attribute

• arp_ip_target: Set bond attribute

• arp_validate: Set bond attribute

• downdelay: Set bond attribute

• fail_over_mac: Set bond attribute

• lacp_rate: Set bond attribute

• lp_interval: Set bond attribute

• miimon: Set bond attribute

• mii_status: Set bond attribute

• min_links: Set bond attribute

• mode: Set bond attribute

• num_grat_arp: Set bond attribute

• num_unsol_na: Set bond attribute

• packets_per_slave: Set bond attribute

• primary: Set bond attribute

• primary_reselect: Set bond attribute

• queue_id: Set bond attribute

• resend_igmp: Set bond attribute

• slaves: Set bond attribute

• tlb_dynamic_lb: Set bond attribute

• updelay: Set bond attribute

• use_carrier: Set bond attribute

• xmit_hash_policy: Set bond attribute

Bond Example:

network:
 version: 1
 config:
   # Simple network adapter
   - type: physical
     name: interface0
     mac_address: '00:11:22:33:44:55'
   # 10G pair
   - type: physical
     name: gbe0
     mac_address: cd:11:22:33:44:00
   - type: physical
     name: gbe1
     mac_address: cd:11:22:33:44:02
   - type: bond
     name: bond0
     bond_interfaces:
       - gbe0
       - gbe1
     params:
       bond-mode: active-backup

Bridge

Type bridge requires the following keys:

• name: Set the name of the bridge.

• bridge_interfaces: Specify the ports of a bridge via their name. This list may be empty.

• params: A list of bridge params. For more details, please read the bridge-utils-interfaces manpage.

Valid keys are:

• bridge_ageing: Set the bridge’s ageing value.

• bridge_bridgeprio: Set the bridge device network priority.

• bridge_fd: Set the bridge’s forward delay.

• bridge_hello: Set the bridge’s hello value.

• bridge_hw: Set the bridge’s MAC address.

• bridge_maxage: Set the bridge’s maxage value.

• bridge_maxwait: Set how long network scripts should wait for the bridge to be up.

• bridge_pathcost: Set the cost of a specific port on the bridge.

• bridge_portprio: Set the priority of a specific port on the bridge.

• bridge_ports: List of devices that are part of the bridge.

• bridge_stp: Set spanning tree protocol on or off.

• bridge_waitport: Set amount of time in seconds to wait on specific ports to become available.

Bridge Example:

network:
 version: 1
 config:
   # Simple network adapter
   - type: physical
     name: interface0
     mac_address: '00:11:22:33:44:55'
   # Second nic with Jumbo frames
   - type: physical
     name: jumbo0
     mac_address: aa:11:22:33:44:55
     mtu: 9000
   - type: bridge
     name: br0
     bridge_interfaces:
       - jumbo0
     params:
       bridge_ageing: 250
       bridge_bridgeprio: 22
       bridge_fd: 1
       bridge_hello: 1
       bridge_maxage: 10
       bridge_maxwait: 0
       bridge_pathcost:
         - jumbo0 75
       bridge_pathprio:
         - jumbo0 28
       bridge_stp: 'off'
       bridge_maxwait:
         - jumbo0 0

VLAN

Type vlan requires the following keys:

• name: Set the name of the VLAN

• vlan_link: Specify the underlying link via its name.

• vlan_id: Specify the VLAN numeric id.

The following optional keys are supported:

mtu: <MTU SizeBytes>

The MTU key represents a device’s Maximum Transmission Unit, the largest size packet or frame, specified in octets (eight-bit bytes), that can be sent in a packet- or frame-based network. Specifying mtu is optional.

Note

The possible supported values of a device’s MTU is not available at configuration time. It’s possible to specify a value too large or to small for a device and may be ignored by the device.

VLAN Example:

network:
  version: 1
  config:
    # Physical interfaces.
    - type: physical
      name: eth0
      mac_address: c0:d6:9f:2c:e8:80
    # VLAN interface.
    - type: vlan
      name: eth0.101
      vlan_link: eth0
      vlan_id: 101
      mtu: 1500

Nameserver

Users can specify a nameserver type. Nameserver dictionaries include the following keys:

• address: List of IPv4 or IPv6 address of nameservers.

• search: List of hostnames to include in the resolv.conf search path.

• interface: Optional. Ties the nameserver definition to the specified interface. The value specified here must match the name of an interface defined in this config. If unspecified, this nameserver will be considered a global nameserver.

Nameserver Example:

network:
  version: 1
  config:
    - type: physical
      name: interface0
      mac_address: '00:11:22:33:44:55'
      subnets:
         - type: static
           address: 192.168.23.14/27
           gateway: 192.168.23.1
    - type: nameserver
      interface: interface0  # Ties nameserver to interface0 only
      address:
        - 192.168.23.2
        - 8.8.8.8
      search:
        - exemplary

Route

Users can include static routing information as well. A route dictionary has the following keys:

• destination: IPv4 network address with CIDR netmask notation.

• gateway: IPv4 gateway address with CIDR netmask notation.

• metric: Integer which sets the network metric value for this route.

Route Example:

network:
  version: 1
  config:
    - type: physical
      name: interface0
      mac_address: '00:11:22:33:44:55'
      subnets:
         - type: static
           address: 192.168.23.14/24
           gateway: 192.168.23.1
    - type: route
      destination: 192.168.24.0/24
      gateway: 192.168.24.1
      metric: 3

Subnet/IP

For any network device (one of the Config Types) users can define a list of subnets which contain ip configuration dictionaries. Multiple subnet entries will create interface alias allowing a single interface to use different ip configurations.

Valid keys for subnets include the following:

• type: Specify the subnet type.

• control: Specify manual, auto or hotplug. Indicates how the interface will be handled during boot.

• address: IPv4 or IPv6 address. It may include CIDR netmask notation.

• netmask: IPv4 subnet mask in dotted format or CIDR notation.

• gateway: IPv4 address of the default gateway for this subnet.

• dns_nameservers: Specify a list of IPv4 dns server IPs to end up in resolv.conf.

• dns_search: Specify a list of search paths to be included in resolv.conf.

• routes: Specify a list of routes for a given interface

Subnet types are one of the following:

• dhcp4: Configure this interface with IPv4 dhcp.

• dhcp: Alias for dhcp4

• dhcp6: Configure this interface with IPv6 dhcp.

• static: Configure this interface with a static IPv4.

• static6: Configure this interface with a static IPv6 .

• ipv6_dhcpv6-stateful: Configure this interface with dhcp6

• ipv6_dhcpv6-stateless: Configure this interface with SLAAC and DHCP

• ipv6_slaac: Configure address with SLAAC

When making use of dhcp or either of the ipv6_dhcpv6 types, no additional configuration is needed in the subnet dictionary.

Using ipv6_dhcpv6-stateless or ipv6_slaac allows the IPv6 address to be automatically configured with StateLess Address AutoConfiguration (SLAAC). SLAAC requires support from the network, so verify that your cloud or network offering has support before trying it out. With ipv6_dhcpv6-stateless, DHCPv6 is still used to fetch other subnet details such as gateway or DNS servers. If you only want to discover the address, use ipv6_slaac.

Subnet DHCP Example:

network:
  version: 1
  config:
    - type: physical
      name: interface0
      mac_address: '00:11:22:33:44:55'
      subnets:
        - type: dhcp

Subnet Static Example:

network:
  version: 1
  config:
    - type: physical
      name: interface0
      mac_address: '00:11:22:33:44:55'
      subnets:
        - type: static
          address: 192.168.23.14/27
          gateway: 192.168.23.1
          dns_nameservers:
            - 192.168.23.2
            - 8.8.8.8
          dns_search:
            - exemplary.maas

The following will result in an interface0 using DHCP and interface0:1 using the static subnet configuration.

Multiple subnet Example:

network:
  version: 1
  config:
    - type: physical
      name: interface0
      mac_address: '00:11:22:33:44:55'
      subnets:
        - type: dhcp
        - type: static
          address: 192.168.23.14/27
          gateway: 192.168.23.1
          dns_nameservers:
            - 192.168.23.2
            - 8.8.8.8
          dns_search:
            - exemplary

Subnet with routes Example:

network:
  version: 1
  config:
    - type: physical
      name: interface0
      mac_address: '00:11:22:33:44:55'
      subnets:
        - type: dhcp
        - type: static
          address: 10.184.225.122
          netmask: 255.255.255.252
          routes:
            - gateway: 10.184.225.121
              netmask: 255.240.0.0
              network: 10.176.0.0
            - gateway: 10.184.225.121
              netmask: 255.240.0.0
              network: 10.208.0.0

Multi-layered configurations

Complex networking sometimes uses layers of configuration. The syntax allows users to build those layers one at a time. All of the virtual network devices supported allow specifying an underlying device by their name value.

Bonded VLAN Example:

network:
  version: 1
  config:
    # 10G pair
    - type: physical
      name: gbe0
      mac_address: cd:11:22:33:44:00
    - type: physical
      name: gbe1
      mac_address: cd:11:22:33:44:02
    # Bond.
    - type: bond
      name: bond0
      bond_interfaces:
        - gbe0
        - gbe1
      params:
        bond-mode: 802.3ad
        bond-lacp-rate: fast
    # A Bond VLAN.
    - type: vlan
        name: bond0.200
        vlan_link: bond0
        vlan_id: 200
        subnets:
            - type: dhcp4

More Examples

Some more examples to explore the various options available.

Multiple VLAN example:

network:
  version: 1
  config:
  - id: eth0
    mac_address: d4:be:d9:a8:49:13
    mtu: 1500
    name: eth0
    subnets:
    - address: 10.245.168.16/21
      dns_nameservers:
      - 10.245.168.2
      gateway: 10.245.168.1
      type: static
    type: physical
  - id: eth1
    mac_address: d4:be:d9:a8:49:15
    mtu: 1500
    name: eth1
    subnets:
    - address: 10.245.188.2/24
      dns_nameservers: []
      type: static
    type: physical
  - id: eth1.2667
    mtu: 1500
    name: eth1.2667
    subnets:
    - address: 10.245.184.2/24
      dns_nameservers: []
      type: static
    type: vlan
    vlan_id: 2667
    vlan_link: eth1
  - id: eth1.2668
    mtu: 1500
    name: eth1.2668
    subnets:
    - address: 10.245.185.1/24
      dns_nameservers: []
      type: static
    type: vlan
    vlan_id: 2668
    vlan_link: eth1
  - id: eth1.2669
    mtu: 1500
    name: eth1.2669
    subnets:
    - address: 10.245.186.1/24
      dns_nameservers: []
      type: static
    type: vlan
    vlan_id: 2669
    vlan_link: eth1
  - id: eth1.2670
    mtu: 1500
    name: eth1.2670
    subnets:
    - address: 10.245.187.2/24
      dns_nameservers: []
      type: static
    type: vlan
    vlan_id: 2670
    vlan_link: eth1
  - address: 10.245.168.2
    search:
    - dellstack
    type: nameserver

Networking Config Version 2

Cloud-init’s support for Version 2 network config is a subset of the version 2 format defined for the netplan tool. Cloud-init supports both reading and writing of Version 2; the latter support requires a distro with netplan present.

Netplan Passthrough

On a system with netplan present, cloud-init will pass Version 2 configuration through to netplan without modification. On such systems, you do not need to limit yourself to the below subset of netplan’s configuration format.

Warning

If you are writing or generating network configuration that may be used on non-netplan systems, you must limit yourself to the subset described in this document, or you will see network configuration failures on non-netplan systems.

Version 2 Configuration Format

The network key has at least two required elements. First it must include version: 2 and one or more of possible device types.

Cloud-init will read this format from system config. For example the following could be present in /etc/cloud/cloud.cfg.d/custom-networking.cfg:

network:
  version: 2
  ethernets: []

It may also be provided in other locations including the NoCloud, see Default Behavior for other places.

Supported device types values are as follows:

• Ethernets (ethernets)

• Bonds (bonds)

• Bridges (bridges)

• VLANs (vlans)

Each type block contains device definitions as a map where the keys (called “configuration IDs”). Each entry under the types may include IP and/or device configuration.

Device configuration IDs

The key names below the per-device-type definition maps (like ethernets:) are called “ID”s. They must be unique throughout the entire set of configuration files. Their primary purpose is to serve as anchor names for composite devices, for example to enumerate the members of a bridge that is currently being defined.

There are two physically/structurally different classes of device definitions, and the ID field has a different interpretation for each:

Physical devices (Examples: ethernet, wifi):

These can dynamically come and go between reboots and even during runtime (hotplugging). In the generic case, they can be selected by match: rules on desired properties, such as name/name pattern, MAC address, driver, or device paths. In general these will match any number of devices (unless they refer to properties which are unique such as the full path or MAC address), so without further knowledge about the hardware these will always be considered as a group.

It is valid to specify no match rules at all, in which case the ID field is simply the interface name to be matched. This is mostly useful if you want to keep simple cases simple, and it’s how network device configuration has been done for a long time.

If there are match: rules, then the ID field is a purely opaque name which is only being used for references from definitions of compound devices in the config.

Virtual devices (Examples: veth, bridge, bond):

These are fully under the control of the config file(s) and the network stack. I. e. these devices are being created instead of matched. Thus match: and set-name: are not applicable for these, and the ID field is the name of the created virtual device.

Common properties for physical device types

match: <(mapping)>

This selects a subset of available physical devices by various hardware properties. The following configuration will then apply to all matching devices, as soon as they appear. All specified properties must match. The following properties for creating matches are supported:

name: <(scalar)>

Current interface name. Globs are supported, and the primary use case for matching on names, as selecting one fixed name can be more easily achieved with having no match: at all and just using the ID (see above). Note that currently only networkd supports globbing, NetworkManager does not.

macaddress: <(scalar)>

Device’s MAC address in the form xx:xx:xx:xx:xx:xx. Globs are not allowed. Letters must be lowercase.

Note

MAC addresses must be strings. As MAC addresses which consist of only the digits 0-9 (i.e. no hex a-f) can be interpreted as a base 60 integer per the YAML 1.1 spec it is best practice to quote all MAC addresses to ensure they are parsed as strings regardless of value.

driver: <(scalar)>

Kernel driver name, corresponding to the DRIVER udev property. Globs are supported. Matching on driver is only supported with networkd.

Examples:

# all cards on second PCI bus
match:
  name: enp2*

# fixed MAC address
match:
  macaddress: 11:22:33:aa:bb:ff

# first card of driver ``ixgbe``
match:
  driver: ixgbe
  name: en*s0

set-name: <(scalar)>

When matching on unique properties such as path or MAC, or with additional assumptions such as “there will only ever be one wifi device”, match rules can be written so that they only match one device. Then this property can be used to give that device a more specific/desirable/nicer name than the default from udev’s ifnames. Any additional device that satisfies the match rules will then fail to get renamed and keep the original kernel name (and dmesg will show an error).

wakeonlan: <(bool)>

Enable wake on LAN. Off by default.

Common properties for all device types

renderer: <(scalar)>

Use the given networking backend for this definition. Currently supported are networkd and NetworkManager. This property can be specified globally in networks:, for a device type (in e. g. ethernets:) or for a particular device definition. Default is networkd.

Note

Cloud-init only supports networkd backend if rendering version2 config to the instance.

dhcp4: <(bool)>

Enable DHCP for IPv4. Off by default.

dhcp6: <(bool)>

Enable DHCP for IPv6. Off by default.

addresses: <(sequence of scalars)>

Add static addresses to the interface in addition to the ones received through DHCP or RA. Each sequence entry is in CIDR notation, i. e. of the form addr/prefixlen . addr is an IPv4 or IPv6 address as recognized by inet_pton``(3) and ``prefixlen the number of bits of the subnet.

Example: addresses: [192.168.14.2/24, 2001:1::1/64]

gateway4: or gateway6: <(scalar)>

Set default gateway for IPv4/6, for manual address configuration. This requires setting addresses too. Gateway IPs must be in a form recognized by inet_pton(3)

Example for IPv4: gateway4: 172.16.0.1 Example for IPv6: gateway6: 2001:4::1

mtu: <MTU SizeBytes>

The MTU key represents a device’s Maximum Transmission Unit, the largest size packet or frame, specified in octets (eight-bit bytes), that can be sent in a packet- or frame-based network. Specifying mtu is optional.

nameservers: <(mapping)>

Set DNS servers and search domains, for manual address configuration. There are two supported fields: addresses: is a list of IPv4 or IPv6 addresses similar to gateway*, and search: is a list of search domains.

Example:

nameservers:
  search: [lab, home]
  addresses: [8.8.8.8, FEDC::1]

routes: <(sequence of mapping)>

Add device specific routes. Each mapping includes a to, via key with an IPv4 or IPv6 address as value. metric is an optional value.

Example:

routes:
 - to: 0.0.0.0/0
   via: 10.23.2.1
   metric: 3

Ethernets

Ethernet device definitions do not support any specific properties beyond the common ones described above.

Bonds

interfaces <(sequence of scalars)>

All devices matching this ID list will be added to the bond.

Example:

ethernets:
  switchports:
    match: {name: "enp2*"}
[...]
bonds:
  bond0:
    interfaces: [switchports]

parameters: <(mapping)>

Customization parameters for special bonding options. Time values are specified in seconds unless otherwise specified.

mode: <(scalar)>

Set the bonding mode used for the interfaces. The default is balance-rr (round robin). Possible values are balance-rr, active-backup, balance-xor, broadcast, 802.3ad, balance-tlb, and balance-alb.

lacp-rate: <(scalar)>

Set the rate at which LACPDUs are transmitted. This is only useful in 802.3ad mode. Possible values are slow (30 seconds, default), and fast (every second).

mii-monitor-interval: <(scalar)>

Specifies the interval for MII monitoring (verifying if an interface of the bond has carrier). The default is 0; which disables MII monitoring.

min-links: <(scalar)>

The minimum number of links up in a bond to consider the bond interface to be up.

transmit-hash-policy: <(scalar)>

Specifies the transmit hash policy for the selection of slaves. This is only useful in balance-xor, 802.3ad and balance-tlb modes. Possible values are layer2, layer3+4, layer2+3, encap2+3, and encap3+4.

ad-select: <(scalar)>

Set the aggregation selection mode. Possible values are stable, bandwidth, and count. This option is only used in 802.3ad mode.

all-slaves-active: <(bool)>

If the bond should drop duplicate frames received on inactive ports, set this option to false. If they should be delivered, set this option to true. The default value is false, and is the desirable behavior in most situations.

arp-interval: <(scalar)>

Set the interval value for how frequently ARP link monitoring should happen. The default value is 0, which disables ARP monitoring.

arp-ip-targets: <(sequence of scalars)>

IPs of other hosts on the link which should be sent ARP requests in order to validate that a slave is up. This option is only used when arp-interval is set to a value other than 0. At least one IP address must be given for ARP link monitoring to function. Only IPv4 addresses are supported. You can specify up to 16 IP addresses. The default value is an empty list.

arp-validate: <(scalar)>

Configure how ARP replies are to be validated when using ARP link monitoring. Possible values are none, active, backup, and all.

arp-all-targets: <(scalar)>

Specify whether to use any ARP IP target being up as sufficient for a slave to be considered up; or if all the targets must be up. This is only used for active-backup mode when arp-validate is enabled. Possible values are any and all.

up-delay: <(scalar)>

Specify the delay before enabling a link once the link is physically up. The default value is 0.

down-delay: <(scalar)>

Specify the delay before disabling a link once the link has been lost. The default value is 0.

fail-over-mac-policy: <(scalar)>

Set whether to set all slaves to the same MAC address when adding them to the bond, or how else the system should handle MAC addresses. The possible values are none, active, and follow.

gratuitious-arp: <(scalar)>

Specify how many ARP packets to send after failover. Once a link is up on a new slave, a notification is sent and possibly repeated if this value is set to a number greater than 1. The default value is 1 and valid values are between 1 and 255. This only affects active-backup mode.

packets-per-slave: <(scalar)>

In balance-rr mode, specifies the number of packets to transmit on a slave before switching to the next. When this value is set to 0, slaves are chosen at random. Allowable values are between 0 and 65535. The default value is 1. This setting is only used in balance-rr mode.

primary-reselect-policy: <(scalar)>

Set the reselection policy for the primary slave. On failure of the active slave, the system will use this policy to decide how the new active slave will be chosen and how recovery will be handled. The possible values are always, better, and failure.

learn-packet-interval: <(scalar)>

Specify the interval between sending learning packets to each slave. The value range is between 1 and 0x7fffffff. The default value is 1. This option only affects balance-tlb and balance-alb modes.

Bridges

interfaces: <(sequence of scalars)>

All devices matching this ID list will be added to the bridge.

Example:

ethernets:
  switchports:
    match: {name: "enp2*"}
[...]
bridges:
  br0:
    interfaces: [switchports]

parameters: <(mapping)>

Customization parameters for special bridging options. Time values are specified in seconds unless otherwise specified.

ageing-time: <(scalar)>

Set the period of time to keep a MAC address in the forwarding database after a packet is received.

priority: <(scalar)>

Set the priority value for the bridge. This value should be an number between 0 and 65535. Lower values mean higher priority. The bridge with the higher priority will be elected as the root bridge.

forward-delay: <(scalar)>

Specify the period of time the bridge will remain in Listening and Learning states before getting to the Forwarding state. This value should be set in seconds for the systemd backend, and in milliseconds for the NetworkManager backend.

hello-time: <(scalar)>

Specify the interval between two hello packets being sent out from the root and designated bridges. Hello packets communicate information about the network topology.

max-age: <(scalar)>

Set the maximum age of a hello packet. If the last hello packet is older than that value, the bridge will attempt to become the root bridge.

path-cost: <(scalar)>

Set the cost of a path on the bridge. Faster interfaces should have a lower cost. This allows a finer control on the network topology so that the fastest paths are available whenever possible.

stp: <(bool)>

Define whether the bridge should use Spanning Tree Protocol. The default value is “true”, which means that Spanning Tree should be used.

VLANs

id: <(scalar)>

VLAN ID, a number between 0 and 4094.

link: <(scalar)>

ID of the underlying device definition on which this VLAN gets created.

Example:

ethernets:
  eno1: {...}
vlans:
  en-intra:
    id: 1
    link: eno1
    dhcp4: yes
  en-vpn:
    id: 2
    link: eno1
    address: ...

Examples

Configure an ethernet device with networkd, identified by its name, and enable DHCP:

network:
  version: 2
  ethernets:
    eno1:
      dhcp4: true

This is a complex example which shows most available features:

network:
  version: 2
  ethernets:
    # opaque ID for physical interfaces, only referred to by other stanzas
    id0:
      match:
        macaddress: '00:11:22:33:44:55'
      wakeonlan: true
      dhcp4: true
      addresses:
        - 192.168.14.2/24
        - 2001:1::1/64
      gateway4: 192.168.14.1
      gateway6: 2001:1::2
      nameservers:
        search: [foo.local, bar.local]
        addresses: [8.8.8.8]
      # static routes
      routes:
        - to: 192.0.2.0/24
          via: 11.0.0.1
          metric: 3
    lom:
      match:
        driver: ixgbe
      # you are responsible for setting tight enough match rules
      # that only match one device if you use set-name
      set-name: lom1
      dhcp6: true
    switchports:
      # all cards on second PCI bus; unconfigured by themselves, will be added
      # to br0 below
      match:
        name: enp2*
      mtu: 1280
  bonds:
    bond0:
      interfaces: [id0, lom]
  bridges:
    # the key name is the name for virtual (created) interfaces; no match: and
    # set-name: allowed
    br0:
      # IDs of the components; switchports expands into multiple interfaces
      interfaces: [wlp1s0, switchports]
      dhcp4: true
  vlans:
    en-intra:
      id: 1
      link: id0
      dhcp4: yes

Network Configuration Outputs

Cloud-init converts various forms of user supplied or automatically generated configuration into an internal network configuration state. From this state Cloud-init delegates rendering of the configuration to Distro supported formats. The following renderers are supported in cloud-init:

• ENI

/etc/network/interfaces or ENI is supported by the ifupdown package found in Alpine Linux, Debian and Ubuntu.

• Netplan

Introduced in Ubuntu 16.10 (Yakkety Yak), netplan has been the default network configuration tool in Ubuntu since 17.10 (Artful Aardvark). netplan consumes Networking Config Version 2 input and renders network configuration for supported backends such as systemd-networkd and NetworkManager.

• Sysconfig

Sysconfig format is used by RHEL, CentOS, Fedora and other derivatives.

Network Output Policy

The default policy for selecting a network renderer in order of preference is as follows:

• ENI

• Sysconfig

• Netplan

When applying the policy, Cloud-init checks if the current instance has the correct binaries and paths to support the renderer. The first renderer that can be used is selected. Users may override the network renderer policy by supplying an updated configuration in cloud-config.

system_info:
  network:
    renderers: ['netplan', 'eni', 'sysconfig', 'freebsd', 'netbsd', 'openbsd']

Network Configuration Tools

Cloud-init contains one tool used to test input/output conversion between formats. The tools/net-convert.py in the Cloud-init source repository is helpful for examining expected output for a given input format.

CLI Interface :

% tools/net-convert.py --help
usage: net-convert.py [-h] --network-data PATH --kind
                      {eni,network_data.json,yaml} -d PATH [-m name,mac]
                      --output-kind {eni,netplan,sysconfig}

optional arguments:
  -h, --help            show this help message and exit
  --network-data PATH, -p PATH
  --kind {eni,network_data.json,yaml}, -k {eni,network_data.json,yaml}
  -d PATH, --directory PATH
                        directory to place output in
  -m name,mac, --mac name,mac
                        interface name to mac mapping
  --output-kind {eni,netplan,sysconfig}, -ok {eni,netplan,sysconfig}

Example output converting V2 to sysconfig:

% tools/net-convert.py --network-data v2.yaml --kind yaml \
    --output-kind sysconfig -d target
% cat target/etc/sysconfig/network-scripts/ifcfg-eth*
# Created by cloud-init on instance boot automatically, do not edit.
#
BOOTPROTO=static
DEVICE=eth7
IPADDR=192.168.1.5/255.255.255.0
NM_CONTROLLED=no
ONBOOT=yes
TYPE=Ethernet
USERCTL=no
# Created by cloud-init on instance boot automatically, do not edit.
#
BOOTPROTO=dhcp
DEVICE=eth9
NM_CONTROLLED=no
ONBOOT=yes
TYPE=Ethernet
USERCTL=no

Contributing to cloud-init

This document describes how to contribute changes to cloud-init. It assumes you have a GitHub account, and refers to your GitHub user as GH_USER throughout.

Submitting your first pull request

Summary

Before any pull request can be accepted, you must do the following:

• Sign the Canonical contributor license agreement

• Add yourself (alphabetically) to the in-repository list that we use to track CLA signatures: tools/.github-cla-signers

• Add or update any unit tests accordingly

• Add or update any integration tests (if applicable)

• Format code (using black and isort) with tox -e do_format

• Ensure unit tests and linting pass using tox

• Submit a PR against the main branch of the cloud-init repository

The detailed instructions

Follow these steps to submit your first pull request to cloud-init:

• To contribute to cloud-init, you must sign the Canonical contributor license agreement

  • If you have already signed it as an individual, your Launchpad user will be listed in the contributor-agreement-canonical group. (Unfortunately there is no easy way to check if an organization or company you are doing work for has signed.)

  • When signing it:

    • ensure that you fill in the GitHub username field.

    • when prompted for ‘Project contact’ or ‘Canonical Project Manager’, enter ‘James Falcon’.

  • If your company has signed the CLA for you, please contact us to help in verifying which Launchpad/GitHub accounts are associated with the company.

  • For any questions or help with the process, please email James Falcon with the subject, “Cloud-Init CLA”

  • You also may contact user falcojr in the #cloud-init channel on the Libera IRC network.

• Configure git with your email and name for commit messages.

  Your name will appear in commit messages and will also be used in changelogs or release notes. Give yourself credit!:

  git config user.name "Your Name"
  git config user.email "Your Email"
  

• Sign into your GitHub account

• Fork the upstream repository on Github and clicking on the Fork button

• Create a new remote pointing to your personal GitHub repository.

  git clone git://github.com/canonical/cloud-init
  cd cloud-init
  git remote add GH_USER git@github.com:GH_USER/cloud-init.git
  git push GH_USER main
  

• Read through the cloud-init Code Review Process, so you understand how your changes will end up in cloud-init’s codebase.

• Submit your first cloud-init pull request, adding yourself to the in-repository list that we use to track CLA signatures: tools/.github-cla-signers

  • See PR #344 and PR #345 for examples of what this pull request should look like.

  • Note that .github-cla-signers is sorted alphabetically.

  • (If you already have a change that you want to submit, you can also include the change to tools/.github-cla-signers in that pull request, there is no need for two separate PRs.)

Transferring CLA Signatures from Launchpad to Github

For existing contributors who have signed the agreement in Launchpad before the Github username field was included, we need to verify the link between your Launchpad account and your GitHub account. To enable us to do this, we ask that you create a branch with both your Launchpad and GitHub usernames against both the Launchpad and GitHub cloud-init repositories. We’ve added a tool (tools/migrate-lp-user-to-github) to the cloud-init repository to handle this migration as automatically as possible.

The cloud-init team will review the two merge proposals and verify that the CLA has been signed for the Launchpad user and record the associated GitHub account.

Note

If you are a first time contributor, you will not need to touch Launchpad to contribute to cloud-init: all new CLA signatures are handled as part of the GitHub pull request process described above.

Do these things for each feature or bug

• Create a new topic branch for your work:

  git checkout -b my-topic-branch
  

• Make and commit your changes (note, you can make multiple commits, fixes, more commits.):

  git commit
  

• Apply black and isort formatting rules with tox:

  tox -e format
  

• Run unit tests and lint/formatting checks with tox:

  tox
  

• Push your changes to your personal GitHub repository:

  git push -u GH_USER my-topic-branch
  

• Use your browser to create a pull request:

  • Open the branch on GitHub

    • You can see a web view of your repository and navigate to the branch at:

      https://github.com/GH_USER/cloud-init/tree/my-topic-branch

  • Click ‘Pull Request`

  • Fill out the pull request title, summarizing the change and a longer message indicating important details about the changes included, like

    Activate the frobnicator.
    
    The frobnicator was previously inactive and now runs by default.
    This may save the world some day.  Then, list the bugs you fixed
    as footers with syntax as shown here.
    
    The commit message should be one summary line of less than
    70 characters followed by a blank line, and then one or more
    paragraphs wrapped at 72 characters describing the change and why
    it was needed.
    
    This is the message that will be used on the commit when it
    is sqaushed and merged into main. If there is a related launchpad
    bug, specify it at the bottom of the commit message.
    
    LP: #NNNNNNN (replace with the appropriate bug reference or remove
    this line entirely if there is no associated bug)
    

    Note that the project continues to use LP: #NNNNN format for closing launchpad bugs rather than GitHub Issues.

  • Click ‘Create Pull Request`

Then, a cloud-init committer will review your changes and follow up in the pull request. Look at the Code Review Process doc to understand the following steps.

Feel free to ping and/or join #cloud-init on Libera irc if you have any questions.

Design

This section captures design decisions that are helpful to know when hacking on cloud-init.

Python Support

Cloud-init upstream currently supports Python 3.6 and above.

Cloud-init upstream will stay compatible with a particular python version for 6 years after release. After 6 years, we will stop testing upstream changes against the unsupported version of python and may introduce breaking changes. This policy may change as needed.

The following table lists the cloud-init versions in which the minimum python version changed:

Cloud-init version	Python version
22.1	3.6+
20.3	3.5+
19.4	2.7+

Cloud Config Modules

• Any new modules should use underscores in any new config options and not hyphens (e.g. new_option and not new-option).

Tests

Submissions to cloud-init must include testing. See Testing for details on these requirements.

Type Annotations

The cloud-init codebase uses Python’s annotation support for storing type annotations in the style specified by PEP-484 and PEP-526. Their use in the codebase is encouraged.

Feature Flags

Feature flags are used as a way to easily toggle configuration at build time. They are provided to accommodate feature deprecation and downstream configuration changes.

Currently used upstream values for feature flags are set in cloudinit/features.py. Overrides to these values (typically via quilt patch) can be placed in a file called feature_overrides.py in the same directory. Any value set in feature_overrides.py will override the original value set in features.py.

Each flag should include a short comment regarding the reason for the flag and intended lifetime.

Tests are required for new feature flags, and tests must verify all valid states of a flag, not just the default state.

cloudinit.features.ALLOW_EC2_MIRRORS_ON_NON_AWS_INSTANCE_TYPES = False

When configuring apt mirrors, if ALLOW_EC2_MIRRORS_ON_NON_AWS_INSTANCE_TYPES is True cloud-init will detect that a datasource’s availability_zone property looks like an EC2 availability zone and set the ec2_region variable when generating mirror URLs; this can lead to incorrect mirrors being configured in clouds whose AZs follow EC2’s naming pattern.

As of 20.3, ALLOW_EC2_MIRRORS_ON_NON_AWS_INSTANCE_TYPES is False so we no longer include ec2_region in mirror determination on non-AWS cloud platforms.

If the old behavior is desired, users can provide the appropriate mirrors via apt: directives in cloud-config.

cloudinit.features.ERROR_ON_USER_DATA_FAILURE = True

If there is a failure in obtaining user data (i.e., #include or decompress fails) and ERROR_ON_USER_DATA_FAILURE is False, cloud-init will log a warning and proceed. If it is True, cloud-init will instead raise an exception.

As of 20.3, ERROR_ON_USER_DATA_FAILURE is True.

(This flag can be removed after Focal is no longer supported.)

Ongoing Refactors

This captures ongoing refactoring projects in the codebase. This is intended as documentation for developers involved in the refactoring, but also for other developers who may interact with the code being refactored in the meantime.

cloudinit.net -> cloudinit.distros.networking Hierarchy

cloudinit.net was imported from the curtin codebase as a chunk, and then modified enough that it integrated with the rest of the cloud-init codebase. Over the ~4 years since, the fact that it is not fully integrated into the Distro hierarchy has caused several issues.

The common pattern of these problems is that the commands used for networking are different across distributions and operating systems. This has lead to cloudinit.net developing its own “distro determination” logic: get_interfaces_by_mac is probably the clearest example of this. Currently, these differences are primarily split along Linux/BSD lines. However, it would be short-sighted to only refactor in a way that captures this difference: we can anticipate that differences will develop between Linux-based distros in future, or there may already be differences in tooling that we currently work around in less obvious ways.

The high-level plan is to introduce a hierarchy of networking classes in cloudinit.distros.networking, which each Distro subclass will reference. These will capture the differences between networking on our various distros, while still allowing easy reuse of code between distros that share functionality (e.g. most of the Linux networking behaviour). Distro objects will instantiate the networking classes at self.networking, so callers will call distro.networking.<func> instead of cloudinit.net.<func>; this will necessitate access to an instantiated Distro object.

An implementation note: there may be external consumers of the cloudinit.net module. We don’t consider this a public API, so we will be removing it as part of this refactor. However, we will ensure that the new API is complete from its introduction, so that any such consumers can move over to it wholesale. (Note, however, that this new API is still not considered public or stable, and may not replicate the existing API exactly.)

In more detail:

• The root of this hierarchy will be the cloudinit.distros.networking.Networking class. This class will have a corresponding method for every cloudinit.net function that we identify to be involved in refactoring. Initially, these methods’ implementations will simply call the corresponding cloudinit.net function. (This gives us the complete API from day one, for existing consumers.)

• As the biggest differentiator in behaviour, the next layer of the hierarchy will be two subclasses: LinuxNetworking and BSDNetworking. These will be introduced in the initial PR.

• When a difference in behaviour for a particular distro is identified, a new Networking subclass will be created. This new class should generally subclass either LinuxNetworking or BSDNetworking.

• To be clear: Networking subclasses will only be created when needed, we will not create a full hierarchy of per-Distro subclasses up-front.

• Each Distro class will have a class variable (cls.networking_cls) which points at the appropriate networking class (initially this will be either LinuxNetworking or BSDNetworking).

• When Distro classes are instantiated, they will instantiate cls.networking_cls and store the instance at self.networking. (This will be implemented in cloudinit.distros.Distro.__init__.)

• A helper function will be added which will determine the appropriate Distro subclass for the current system, instantiate it and return its networking attribute. (This is the entry point for existing consumers to migrate to.)

• Callers of refactored functions will change from calling cloudinit.net.<func> to distro.networking.<func>, where distro is an instance of the appropriate Distro class for this system. (This will require making such an instance available to callers, which will constitute a large part of the work in this project.)

After the initial structure is in place, the work in this refactor will consist of replacing the cloudinit.net.some_func call in each cloudinit.distros.networking.Networking method with the actual implementation. This can be done incrementally, one function at a time:

• pick an unmigrated cloudinit.distros.networking.Networking method

• find it in the the list of bugs tagged net-refactor and assign yourself to it (see Managing Work/Tracking Progress below for more details)

• refactor all of its callers to call the distro.networking.<func> method on Distro instead of the cloudinit.net.<func> function. (This is likely to be the most time-consuming step, as it may require plumbing Distro objects through to places that previously have not consumed them.)

• refactor its implementation from cloudinit.net into the Networking hierarchy (e.g. if it has an if/else on BSD, this is the time to put the implementations in their respective subclasses)

  • if part of the method contains distro-independent logic, then you may need to create new methods to capture this distro-specific logic; we don’t want to replicate common logic in different Networking subclasses

  • if after the refactor, the method on the root Networking class no longer has any implementation, it should be converted to an abstractmethod

• ensure that the new implementation has unit tests (either by moving existing tests, or by writing new ones)

• ensure that the new implementation has a docstring

• add any appropriate type annotations

  • note that we must follow the constraints described in the “Type Annotations” section above, so you may not be able to write complete annotations

  • we have type aliases defined in cloudinit.distros.networking which should be used when applicable

• finally, remove it (and any other now-unused functions) from cloudinit.net (to avoid having two parallel implementations)

cloudinit.net Functions/Classes

The functions/classes that need refactoring break down into some broad categories:

• helpers for accessing /sys (that should not be on the top-level Networking class as they are Linux-specific):

  • get_sys_class_path

  • sys_dev_path

  • read_sys_net

  • read_sys_net_safe

  • read_sys_net_int

• those that directly access /sys (via helpers) and should (IMO) be included in the API of the Networking class:

  • generate_fallback_config

    • the config_driver parameter is used and passed as a boolean, so we can change the default value to False (instead of None)

  • get_ib_interface_hwaddr

  • get_interface_mac

  • interface_has_own_mac

  • is_bond

  • is_bridge

  • is_physical

  • is_renamed

  • is_up

  • is_vlan

  • wait_for_physdevs

• those that directly access /sys (via helpers) but may be Linux-specific concepts or names:

  • get_master

  • device_devid

  • device_driver

• those that directly use ip:

  • _get_current_rename_info

    • this has non-distro-specific logic so should potentially be refactored to use helpers on self instead of ip directly (rather than being wholesale reimplemented in each of BSDNetworking or LinuxNetworking)

    • we can also remove the check_downable argument, it’s never specified so is always True

  • _rename_interfaces

    • this has several internal helper functions which use ip directly, and it calls _get_current_rename_info. That said, there appears to be a lot of non-distro-specific logic that could live in a function on Networking, so this will require some careful refactoring to avoid duplicating that logic in each of BSDNetworking and LinuxNetworking.

    • only the renames and current_info parameters are ever passed in (and current_info only by tests), so we can remove the others from the definition

  • EphemeralIPv4Network

    • this is another case where it mixes distro-specific and non-specific functionality. Specifically, __init__, __enter__ and __exit__ are non-specific, and the remaining methods are distro-specific.

    • when refactoring this, the need to track cleanup_cmds likely means that the distro-specific behaviour cannot be captured only in the Networking class. See this comment in PR #363 for more thoughts.

• those that implicitly use /sys via their call dependencies:

  • master_is_bridge_or_bond

    • appends to get_master return value, which is a /sys path

  • extract_physdevs

    • calls device_driver and device_devid in both _version_* impls

  • apply_network_config_names

    • calls extract_physdevs

    • there is already a Distro.apply_network_config_names which in the default implementation calls this function; this and its BSD subclass implementations should be refactored at the same time

    • the strict_present and strict_busy parameters are never passed, nor are they used in the function definition, so they can be removed

  • get_interfaces

    • calls device_driver, device_devid amongst others

  • get_ib_hwaddrs_by_interface

    • calls get_interfaces

• those that may fall into the above categories, but whose use is only related to netfailover (which relies on a Linux-specific network driver, so is unlikely to be relevant elsewhere without a substantial refactor; these probably only need implementing in LinuxNetworking):

  • get_dev_features

  • has_netfail_standby_feature

    • calls get_dev_features

  • is_netfailover

  • is_netfail_master

    • this is called from generate_fallback_config

  • is_netfail_primary

  • is_netfail_standby

  • N.B. all of these take an optional driver argument which is used to pass around a value to avoid having to look it up by calling device_driver every time. This is something of a leaky abstraction, and is better served by caching on device_driver or storing the cached value on self, so we can drop the parameter from the new API.

• those that use /sys (via helpers) and have non-exhaustive BSD logic:

  • get_devicelist

• those that already have separate Linux/BSD implementations:

  • find_fallback_nic

  • get_interfaces_by_mac

• those that have no OS-specific functionality (so do not need to be refactored):

  • ParserError

  • RendererNotFoundError

  • has_url_connectivity

  • is_ip_address

  • is_ipv4_address

  • natural_sort_key

Note that the functions in cloudinit.net use inconsistent parameter names for “string that contains a device name”; we can standardise on devname (the most common one) in the refactor.

Managing Work/Tracking Progress

To ensure that we won’t have multiple people working on the same part of the refactor at the same time, there is a bug for each function. You can see the current status by looking at the list of bugs tagged net-refactor.

When you’re working on refactoring a particular method, ensure that you have assigned yourself to the corresponding bug, to avoid duplicate work.

Generally, when considering what to pick up to refactor, it is best to start with functions in cloudinit.net which are not called by anything else in cloudinit.net. This allows you to focus only on refactoring that function and its callsites, rather than having to update the other cloudinit.net function also.

References

• Mina Galić’s email the the cloud-init ML in 2018 (plus its thread)

• Mina Galić’s email to the cloud-init ML in 2019 (plus its thread)

• PR #363, the discussion which prompted finally starting this refactor (and where a lot of the above details were hashed out)

Code Review Process

In order to manage incoming pull requests effectively, and provide timely feedback and/or acceptance this document serves as a guideline for the review process and outlines the expectations for those submitting code to the project as well as those reviewing the code. Code is reviewed for acceptance by at least one core team member (later referred to as committers), but comments and suggestions from others are encouraged and welcome.

The process is intended to provide timely and actionable feedback for any submission.

Asking For Help

cloud-init contributors, potential contributors, community members and users are encouraged to ask for any help that they need. If you have questions about the code review process, or at any point during the code review process, these are the available avenues:

• if you have an open Pull Request, comment on that pull request

• join the #cloud-init channel on the Libera IRC network and ask away

• send an email to the cloud-init mailing list, cloud-init@lists.launchpad.net

These are listed in rough order of preference, but use whichever of them you are most comfortable with.

Goals

This process has the following goals:

• Ensure code reviews occur in a timely fashion and provide actionable feedback if changes are desired.

• Ensure the minimization of ancillary problems to increase the efficiency for those reviewing the submitted code

Role Definitions

Any code review process will have (at least) two involved parties. For our purposes, these parties are referred to as Proposer and Reviewer. (We also have the Committer role which is a special case of the Reviewer role.) The terms are defined here (and the use of the singular form is not meant to imply that they refer to a single person):

Proposer

The person proposing a pull request (hereafter known as a PR).

Reviewer

A person who is reviewing a PR.

Committer

A cloud-init core developer (i.e. a person who has permission to merge PRs into main).

Prerequisites For Landing Pull Requests

Before a PR can be landed into main, the following conditions must be met:

• the CLA has been signed by the Proposer (or is covered by an entity-level CLA signature)

• all required status checks are passing

• at least one “Approve” review from a Committer

• no “Request changes” reviews from any Committer

The following conditions should be met:

• any Python functions/methods/classes have docstrings added/updated

• any changes to config module behaviour are captured in the documentation of the config module

• any Python code added has corresponding unit tests

• no “Request changes” reviews from any Reviewer

These conditions can be relaxed at the discretion of the Committers on a case-by-case basis. Generally, for accountability, this should not be the decision of a single Committer, and the decision should be documented in comments on the PR.

(To take a specific example, the cc_phone_home module had no tests at the time PR #237 was submitted, so the Proposer was not expected to write a full set of tests for their minor modification, but they were expected to update the config module docs.)

Non-Committer Reviews

Reviews from non-Committers are always welcome. Please feel empowered to review PRs and leave your thoughts and comments on any submitted PRs, regardless of the Proposer.

Much of the below process is written in terms of the Committers. This is not intended to reflect that reviews should only come from that group, but acknowledges that we are ultimately responsible for maintaining the standards of the codebase. It would be entirely reasonable (and very welcome) for a Reviewer to only examine part of a PR, but it would not be appropriate for a Committer to merge a PR without full scrutiny.

Opening Phase

In this phase, the Proposer is responsible for opening a pull request and meeting the prerequisites laid out above.

If they need help understanding the prerequisites, or help meeting the prerequisites, then they can (and should!) ask for help. See the Asking For Help section above for the ways to do that.

These are the steps that comprise the opening phase:

1. The Proposer opens PR

2. CI runs automatically, and if

   CI fails

   The Proposer is expected to fix CI failures. If the Proposer doesn’t understand the nature of the failures they are seeing, they should comment in the PR to request assistance, or use another way of Asking For Help.

   (Note that if assistance is not requested, the Committers will assume that the Proposer is working on addressing the failures themselves. If you require assistance, please do ask for help!)

   CI passes

   Move on to the Review Phase.

Review Phase

In this phase, the Proposer and the Reviewers will iterate together to, hopefully, get the PR merged into the cloud-init codebase. There are three potential outcomes: merged, rejected permanently, and temporarily closed. (The first two are covered in this section; see Inactive Pull Requests for details about temporary closure.)

(In the below, when the verbs “merge” or “squash merge” are used, they should be understood to mean “squash merged using the GitHub UI”, which is the only way that changes can land in cloud-init’s main branch.)

These are the steps that comprise the review phase:

1. The Committers assign a Committer to the PR

   This Committer is expected to shepherd the PR to completion (and merge it, if that is the outcome reached). This means that they will perform an initial review, and monitor the PR to ensure that the Proposer is receiving any assistance that they require. The Committers will perform this assignment on a daily basis.

   This assignment is intended to ensure that the Proposer has a clear point of contact with a cloud-init core developer, and that they get timely feedback after submitting a PR. It is not intended to preclude reviews from any other Reviewers, nor to imply that the Committer has ownership over the review process.

   The assigned Committer may choose to delegate the code review of a PR to another Reviewer if they think that they would be better suited.

   (Note that, in GitHub terms, this is setting an Assignee, not requesting a review.)

2. That Committer performs an initial review of the PR, resulting in one of the following:

   Approve

   If the submitted PR meets all of the Prerequisites For Landing Pull Requests and passes code review, then the Committer will squash merge immediately.

   There may be circumstances where a PR should not be merged immediately. The wip label will be applied to PRs for which this is true. Only Committers are able to apply labels to PRs, so anyone who believes that this label should be applied to a PR should request its application in a comment on the PR.

   The review process is DONE.

   Approve (with nits)

   If the Proposer submits their PR with “Allow edits from maintainer” enabled, and the only changes the Committer requests are minor “nits”, the Committer can push fixes for those nits and immediately squash merge. If the Committer does not wish to fix these nits but believes they should block a straight-up Approve, then their review should be “Needs Changes” instead.

   A nit is understood to be something like a minor style issue or a spelling error, generally confined to a single line of code.

   If a Committer is unsure as to whether their requested change is a nit, they should not treat it as a nit.

   (If a Proposer wants to opt-out of this, then they should uncheck “Allow edits from maintainer” when submitting their PR.)

   The review process is DONE.

   Outright rejection

   The Committer will close the PR, with useful messaging for the Proposer as to why this has happened.

   This is reserved for cases where the proposed change is completely unfit for landing, and there is no reasonable path forward. This should only be used sparingly, as there are very few cases where proposals are completely unfit.

   If a different approach to the same problem is planned, it should be submitted as a separate PR. The Committer should include this information in their message when the PR is closed.

   The review process is DONE.

   Needs Changes

   The Committer will give the Proposer a clear idea of what is required for an Approve vote or, for more complex PRs, what the next steps towards an Approve vote are.

   The Proposer will ask questions if they don’t understand, or disagree with, the Committer’s review comments.

   Once consensus has been reached, the Proposer will address the review comments.

   Once the review comments are addressed (as well as, potentially, in the interim), CI will run. If CI fails, the Proposer is expected to fix CI failures. If CI passes, the Proposer should indicate that the PR is ready for re-review (by @ing the assigned reviewer), effectively moving back to the start of this section.

Inactive Pull Requests

PRs will be temporarily closed if they have been waiting on Proposer action for a certain amount of time without activity. A PR will be marked as stale (with an explanatory comment) after 14 days of inactivity. It will be closed after a further 7 days of inactivity.

These closes are not considered permanent, and the closing message should reflect this for the Proposer. However, if a PR is reopened, it should effectively enter the Opening Phase again, as it may need some work done to get CI passing again.

Security

Security Policy

The following documents the upstream cloud-init security policy.

Reporting

If a user finds a security issue, they are requested to file a private security bug on Launchpad. To ensure the information stays private, change the “This bug contains information that is:” from “Public” to “Private Security” when filing.

After the bug is received, the issue is triaged within 2 working days of being reported and a response is sent to the reporter.

cloud-init-security

The cloud-init-security Launchpad team is a private, invite-only team used to discuss and coordinate security issues with the project.

Any issues disclosed to the cloud-init-security mailing list are considered embargoed and should only be discussed with other members of the cloud-init-security mailing list before the coordinated release date, unless specific exception is granted by the administrators of the mailing list. This includes disclosure of any details related to the vulnerability or the presence of a vulnerability itself. Violation of this policy may result in removal from the list for the company or individual involved.

Evaluation

If the reported bug is deemed a real security issue a CVE is assigned by the Canonical Security Team as CVE Numbering Authority (CNA).

If it is deemed a regular, non-security, issue, the reporter will be asked to follow typical bug reporting procedures.

In addition to the disclosure timeline, the core Canonical cloud-init team will enlist the expertise of the Ubuntu Security team for guidance on industry-standard disclosure practices as necessary.

If an issue specifically involves another distro or cloud vendor, additional individuals will be informed of the issue to help in evaluation.

Disclosure

Disclosure of security issues will be made with a public statement. Once the determined time for disclosure has arrived the following will occur:

• A public bug is filed/made public with vulnerability details, CVE, mitigations and where to obtain the fix

• An email is sent to the public cloud-init mailing list

The disclosure timeframe is coordinated with the reporter and members of the cloud-init-security list. This depends on a number of factors:

• The reporter might have their own disclosure timeline (e.g. Google Project Zero and many others use a 90-days after initial report OR when a fix becomes public)

• It might take time to decide upon and develop an appropriate fix

• A distros might want extra time to backport any possible fixes before the fix becomes public

• A cloud may need additional time to prepare to help customers or impliment a fix

• The issue might be deemed low priority

• May wish to to align with an upcoming planned release

Debugging cloud-init

Overview

This topic will discuss general approaches for test and debug of cloud-init on deployed instances.

Boot Time Analysis - cloud-init analyze

Occasionally instances don’t appear as performant as we would like and cloud-init packages a simple facility to inspect what operations took cloud-init the longest during boot and setup.

The script /usr/bin/cloud-init has an analyze sub-command analyze which parses any cloud-init.log file into formatted and sorted events. It allows for detailed analysis of the most costly cloud-init operations are to determine the long-pole in cloud-init configuration and setup. These subcommands default to reading /var/log/cloud-init.log.

• analyze show Parse and organize cloud-init.log events by stage and include each sub-stage granularity with time delta reports.

$ cloud-init analyze show -i my-cloud-init.log
-- Boot Record 01 --
The total time elapsed since completing an event is printed after the "@"
character.
The time the event takes is printed after the "+" character.

Starting stage: modules-config
|`->config-emit_upstart ran successfully @05.47600s +00.00100s
|`->config-snap_config ran successfully @05.47700s +00.00100s
|`->config-ssh-import-id ran successfully @05.47800s +00.00200s
|`->config-locale ran successfully @05.48000s +00.00100s
...

• analyze dump Parse cloud-init.log into event records and return a list of dictionaries that can be consumed for other reporting needs.

$ cloud-init analyze dump -i my-cloud-init.log
[
 {
  "description": "running config modules",
  "event_type": "start",
  "name": "modules-config",
  "origin": "cloudinit",
  "timestamp": 1510807493.0
 },...

• analyze blame Parse cloud-init.log into event records and sort them based on highest time cost for quick assessment of areas of cloud-init that may need improvement.

$ cloud-init analyze blame -i my-cloud-init.log
-- Boot Record 11 --
     00.01300s (modules-final/config-scripts-per-boot)
     00.00400s (modules-final/config-final-message)
     00.00100s (modules-final/config-rightscale_userdata)
     ...

• analyze boot Make subprocess calls to the kernel in order to get relevant pre-cloud-init timestamps, such as the kernel start, kernel finish boot, and cloud-init start.

$ cloud-init analyze boot
-- Most Recent Boot Record --
    Kernel Started at: 2019-06-13 15:59:55.809385
    Kernel ended boot at: 2019-06-13 16:00:00.944740
    Kernel time to boot (seconds): 5.135355
    Cloud-init start: 2019-06-13 16:00:05.738396
    Time between Kernel boot and Cloud-init start (seconds): 4.793656

Analyze quickstart - LXC

To quickly obtain a cloud-init log try using lxc on any ubuntu system:

$ lxc init ubuntu-daily:focal x1
$ lxc start x1
$ # Take lxc's cloud-init.log and pipe it to the analyzer
$ lxc file pull x1/var/log/cloud-init.log - | cloud-init analyze dump -i -
$ lxc file pull x1/var/log/cloud-init.log - | \
  python3 -m cloudinit.analyze dump -i -

Analyze quickstart - KVM

To quickly analyze a KVM a cloud-init log:

1. Download the current cloud image

$ wget https://cloud-images.ubuntu.com/daily/server/focal/current/focal-server-cloudimg-amd64.img

2. Create a snapshot image to preserve the original cloud-image

$ qemu-img create -b focal-server-cloudimg-amd64.img -f qcow2 \
test-cloudinit.qcow2

3. Create a seed image with metadata using cloud-localds

$ cat > user-data <<EOF
  #cloud-config
  password: passw0rd
  chpasswd: { expire: False }
  EOF
$  cloud-localds my-seed.img user-data

4. Launch your modified VM

$  kvm -m 512 -net nic -net user -redir tcp:2222::22 \
    -drive file=test-cloudinit.qcow2,if=virtio,format=qcow2 \
    -drive file=my-seed.img,if=virtio,format=raw

5. Analyze the boot (blame, dump, show)

$ ssh -p 2222 ubuntu@localhost 'cat /var/log/cloud-init.log' | \
   cloud-init analyze blame -i -

Running single cloud config modules

This subcommand is not called by the init system. It can be called manually to load the configured datasource and run a single cloud-config module once using the cached userdata and metadata after the instance has booted. Each cloud-config module has a module FREQUENCY configured: PER_INSTANCE, PER_BOOT, PER_ONCE or PER_ALWAYS. When a module is run by cloud-init, it stores a semaphore file in /var/lib/cloud/instance/sem/config_<module_name>.<frequency> which marks when the module last successfully ran. Presence of this semaphore file prevents a module from running again if it has already been run. To ensure that a module is run again, the desired frequency can be overridden on the commandline:

$ sudo cloud-init single --name cc_ssh --frequency always
...
Generating public/private ed25519 key pair
...

Inspect cloud-init.log for output of what operations were performed as a result.

Stable Release Updates (SRU) testing for cloud-init

Once an Ubuntu release is stable (i.e. after it is released), updates for it must follow a special procedure called a “stable release update” (or SRU).

The cloud-init project has a specific process it follows when validating a cloud-init SRU, documented in the CloudinitUpdates wiki page.

Generally an SRU test of cloud-init performs the following:

• Install a pre-release version of cloud-init from the -proposed APT pocket (e.g. bionic-proposed)

• Upgrade cloud-init and attempt a clean run of cloud-init to assert the new version of cloud-init works properly the specific platform and Ubuntu series

• Check for tracebacks or errors in behavior

Manual SRU verification procedure

Below are steps to manually test a pre-release version of cloud-init from -proposed

Note

For each Ubuntu SRU, the Ubuntu Server team manually validates the new version of cloud-init on these platforms: Amazon EC2, Azure, GCE, OpenStack, Oracle, Softlayer (IBM), LXD, KVM

1. Launch a VM on your favorite platform, providing this cloud-config user-data and replacing <YOUR_LAUNCHPAD_USERNAME> with your username:

## template: jinja
#cloud-config
ssh_import_id: [<YOUR_LAUNCHPAD_USERNAME>]
hostname: SRU-worked-{{v1.cloud_name}}

2. Wait for current cloud-init to complete, replace <YOUR_VM_IP> with the IP address of the VM that you launched in step 1:

CI_VM_IP=<YOUR_VM_IP>
# Make note of the datasource cloud-init detected in --long output.
# In step 5, you will use this to confirm the same datasource is detected after upgrade.
ssh ubuntu@$CI_VM_IP -- cloud-init status --wait --long

3. Set up the -proposed pocket on your VM and upgrade to the -proposed cloud-init:

# Create a script that will add the -proposed pocket to APT's sources
# and install cloud-init from that pocket
cat > setup_proposed.sh <<EOF
#/bin/bash
mirror=http://archive.ubuntu.com/ubuntu
echo deb \$mirror \$(lsb_release -sc)-proposed main | tee \
    /etc/apt/sources.list.d/proposed.list
apt-get update -q
apt-get install -qy cloud-init
EOF

scp setup_proposed.sh ubuntu@$CI_VM_IP:.
ssh ubuntu@$CI_VM_IP -- sudo bash setup_proposed.sh

4. Change hostname, clean cloud-init’s state, and reboot to run cloud-init from scratch:

ssh ubuntu@$CI_VM_IP -- sudo hostname something-else
ssh ubuntu@$CI_VM_IP -- sudo cloud-init clean --logs --reboot

5. Validate -proposed cloud-init came up without error

# Block until cloud-init completes and verify from --long the datasource
# from step 1. Errors would show up in --long

ssh ubuntu@$CI_VM_IP -- cloud-init status --wait --long
# Make sure hostname was set properly to SRU-worked-<cloud name>
ssh ubuntu@$CI_VM_IP -- hostname
# Check for any errors or warnings in cloud-init logs.
# (This should produce no output if successful.)
ssh ubuntu@$CI_VM_IP -- grep Trace "/var/log/cloud-init*"

6. If you encounter an error during SRU testing:

   • Create a new cloud-init bug reporting the version of cloud-init affected

   • Ping upstream cloud-init on Libera’s #cloud-init IRC channel

Logging

Cloud-init supports both local and remote logging configurable through python’s built-in logging configuration and through the cloud-init rsyslog module.

Command Output

Cloud-init can redirect its stdout and stderr based on config given under the output config key. The output of any commands run by cloud-init and any user or vendor scripts provided will also be included here. The output key accepts a dictionary for configuration. Output files may be specified individually for each stage (init, config, and final), or a single key all may be used to specify output for all stages.

The output for each stage may be specified as a dictionary of output and error keys, for stdout and stderr respectively, as a tuple with stdout first and stderr second, or as a single string to use for both. The strings passed to all of these keys are handled by the system shell, so any form of redirection that can be used in bash is valid, including piping cloud-init’s output to tee, or logger. If only a filename is provided, cloud-init will append its output to the file as though >> was specified.

By default, cloud-init loads its output configuration from /etc/cloud/cloud.cfg.d/05_logging.cfg. The default config directs both stdout and stderr from all cloud-init stages to /var/log/cloud-init-output.log. The default config is given as

output: { all: "| tee -a /var/log/cloud-init-output.log" }

For a more complex example, the following configuration would output the init stage to /var/log/cloud-init.out and /var/log/cloud-init.err, for stdout and stderr respectively, replacing anything that was previously there. For the config stage, it would pipe both stdout and stderr through tee -a /var/log/cloud-config.log. For the final stage it would append the output of stdout and stderr to /var/log/cloud-final.out and /var/log/cloud-final.err respectively.

output:
    init:
        output: "> /var/log/cloud-init.out"
        error: "> /var/log/cloud-init.err"
    config: "tee -a /var/log/cloud-config.log"
    final:
        - ">> /var/log/cloud-final.out"
        - "/var/log/cloud-final.err"

Python Logging

Cloud-init uses the python logging module, and can accept config for this module using the standard python fileConfig format. Cloud-init looks for config for the logging module under the logcfg key.

Note

the logging configuration is not YAML, it is python fileConfig format, and is passed through directly to the python logging module. please use the correct syntax for a multi-line string in YAML.

By default, cloud-init uses the logging configuration provided in /etc/cloud/cloud.cfg.d/05_logging.cfg. The default python logging configuration writes all cloud-init events with a priority of WARNING or higher to console, and writes all events with a level of DEBUG or higher to /var/log/cloud-init.log and via syslog.

Python’s fileConfig format consists of sections with headings in the format [title] and key value pairs in each section. Configuration for python logging must contain the sections [loggers], [handlers], and [formatters], which name the entities of their respective types that will be defined. The section name for each defined logger, handler and formatter will start with its type, followed by an underscore (_) and the name of the entity. For example, if a logger was specified with the name log01, config for the logger would be in the section [logger_log01].

Logger config entries contain basic logging set up. They may specify a list of handlers to send logging events to as well as the lowest priority level of events to handle. A logger named root must be specified and its configuration (under [logger_root]) must contain a level and a list of handlers. A level entry can be any of the following: DEBUG, INFO, WARNING, ERROR, CRITICAL, or NOTSET. For the root logger the NOTSET option will allow all logging events to be recorded.

Each configured handler must specify a class under the python’s logging package namespace. A handler may specify a message formatter to use, a priority level, and arguments for the handler class. Common handlers are StreamHandler, which handles stream redirects (i.e. logging to stderr), and FileHandler which outputs to a log file. The logging module also supports logging over net sockets, over http, via smtp, and additional complex configurations. For full details about the handlers available for python logging, please see the documentation for python logging handlers.

Log messages are formatted using the logging.Formatter class, which is configured using formatter config entities. A default format of %(message)s is given if no formatter configs are specified. Formatter config entities accept a format string which supports variable replacements. These may also accept a datefmt string which may be used to configure the timestamp used in the log messages. The format variables %(asctime)s, %(levelname)s and %(message)s are commonly used and represent the timestamp, the priority level of the event and the event message. For additional information on logging formatters see python logging formatters.

Note

by default the format string used in the logging formatter are in python’s old style %s form. the str.format() and string.Template styles can also be used by using { or $ in place of % by setting the style parameter in formatter config.

A simple, but functional python logging configuration for cloud-init is below. It will log all messages of priority DEBUG or higher both stderr and /tmp/my.log using a StreamHandler and a FileHandler, using the default format string %(message)s:

logcfg: |
 [loggers]
 keys=root,cloudinit
 [handlers]
 keys=ch,cf
 [formatters]
 keys=
 [logger_root]
 level=DEBUG
 handlers=
 [logger_cloudinit]
 level=DEBUG
 qualname=cloudinit
 handlers=ch,cf
 [handler_ch]
 class=StreamHandler
 level=DEBUG
 args=(sys.stderr,)
 [handler_cf]
 class=FileHandler
 level=DEBUG
 args=('/tmp/my.log',)

For additional information about configuring python’s logging module, please see the documentation for python logging config.

Rsyslog Module

Cloud-init’s cc_rsyslog module allows for fully customizable rsyslog configuration under the rsyslog config key. The simplest way to use the rsyslog module is by specifying remote servers under the remotes key in rsyslog config. The remotes key takes a dictionary where each key represents the name of an rsyslog server and each value is the configuration for that server. The format for server config is:

• optional filter for log messages (defaults to *.*)

• optional leading @ or @@, indicating udp and tcp respectively (defaults to @, for udp)

• ipv4 or ipv6 hostname or address. ipv6 addresses must be in [::1] format, (e.g. @[fd00::1]:514)

• optional port number (defaults to 514)

For example, to send logging to an rsyslog server named log_serv with address 10.0.4.1, using port number 514, over udp, with all log messages enabled one could use either of the following.

With all options specified:

rsyslog:
    remotes:
        log_serv: "*.* @10.0.4.1:514"

With defaults used:

rsyslog:
    remotes:
        log_serv: "10.0.4.1"

For more information on rsyslog configuration, see Rsyslog.

Directory layout

Cloud-init’s directory structure is somewhat different from a regular application:

/var/lib/cloud/
    - data/
       - instance-id
       - previous-instance-id
       - datasource
       - previous-datasource
       - previous-hostname
    - handlers/
    - instance
    - instances/
        i-00000XYZ/
          - boot-finished
          - cloud-config.txt
          - datasource
          - handlers/
          - obj.pkl
          - scripts/
          - sem/
          - user-data.txt
          - user-data.txt.i
    - scripts/
       - per-boot/
       - per-instance/
       - per-once/
    - seed/
    - sem/

/var/lib/cloud

The main directory containing the cloud-init specific subdirectories. It is typically located at /var/lib but there are certain configuration scenarios where this can be altered.

data/

Contains information related to instance ids, datasources and hostnames of the previous and current instance if they are different. These can be examined as needed to determine any information related to a previous boot (if applicable).

handlers/

Custom part-handlers code is written out here. Files that end up here are written out with in the scheme of part-handler-XYZ where XYZ is the handler number (the first handler found starts at 0).

instance

A symlink to the current instances/ subdirectory that points to the currently active instance (which is active is dependent on the datasource loaded).

instances/

All instances that were created using this image end up with instance identifier subdirectories (and corresponding data for each instance). The currently active instance will be symlinked the instance symlink file defined previously.

scripts/

Scripts that are downloaded/created by the corresponding part-handler will end up in one of these subdirectories.

seed/

Contains seeded data files: meta-data, network-config, user-data, vendor-data

sem/

Cloud-init has a concept of a module semaphore, which basically consists of the module name and its frequency. These files are used to ensure a module is only ran per-once, per-instance, per-always. This folder contains semaphore files which are only supposed to run per-once (not tied to the instance id).

Analyze

The analyze subcommand was added to cloud-init in order to help analyze cloud-init boot time performance. It is loosely based on systemd-analyze where there are four subcommands:

• blame

• show

• dump

• boot

Usage

The analyze command requires one of the four subcommands:

$ cloud-init analyze blame
$ cloud-init analyze show
$ cloud-init analyze dump
$ cloud-init analyze boot

Availability

The analyze subcommand is generally available across all distributions with the exception of Gentoo and FreeBSD.

Subcommands

Blame

The blame action matches systemd-analyze blame where it prints, in descending order, the units that took the longest to run. This output is highly useful for examining where cloud-init is spending its time during execution.

$ cloud-init analyze blame
-- Boot Record 01 --
    00.80300s (init-network/config-growpart)
    00.64300s (init-network/config-resizefs)
    00.62100s (init-network/config-ssh)
    00.57300s (modules-config/config-grub-dpkg)
    00.40300s (init-local/search-NoCloud)
    00.38200s (init-network/config-users-groups)
    00.19800s (modules-config/config-apt-configure)
    00.03700s (modules-final/config-keys-to-console)
    00.02100s (init-network/config-update_etc_hosts)
    00.02100s (init-network/check-cache)
    00.00800s (modules-final/config-ssh-authkey-fingerprints)
    00.00800s (init-network/consume-vendor-data)
    00.00600s (modules-config/config-timezone)
    00.00500s (modules-final/config-final-message)
    00.00400s (init-network/consume-user-data)
    00.00400s (init-network/config-mounts)
    00.00400s (init-network/config-disk_setup)
    00.00400s (init-network/config-bootcmd)
    00.00400s (init-network/activate-datasource)
    00.00300s (init-network/config-update_hostname)
    00.00300s (init-network/config-set_hostname)
    00.00200s (modules-final/config-snappy)
    00.00200s (init-network/config-rsyslog)
    00.00200s (init-network/config-ca-certs)
    00.00200s (init-local/check-cache)
    00.00100s (modules-final/config-scripts-vendor)
    00.00100s (modules-final/config-scripts-per-once)
    00.00100s (modules-final/config-salt-minion)
    00.00100s (modules-final/config-rightscale_userdata)
    00.00100s (modules-final/config-phone-home)
    00.00100s (modules-final/config-package-update-upgrade-install)
    00.00100s (modules-final/config-fan)
    00.00100s (modules-config/config-ubuntu-advantage)
    00.00100s (modules-config/config-ssh-import-id)
    00.00100s (modules-config/config-snap)
    00.00100s (modules-config/config-set-passwords)
    00.00100s (modules-config/config-runcmd)
    00.00100s (modules-config/config-locale)
    00.00100s (modules-config/config-byobu)
    00.00100s (modules-config/config-apt-pipelining)
    00.00100s (init-network/config-write-files)
    00.00100s (init-network/config-seed_random)
    00.00100s (init-network/config-migrator)
    00.00000s (modules-final/config-ubuntu-drivers)
    00.00000s (modules-final/config-scripts-user)
    00.00000s (modules-final/config-scripts-per-instance)
    00.00000s (modules-final/config-scripts-per-boot)
    00.00000s (modules-final/config-puppet)
    00.00000s (modules-final/config-power-state-change)
    00.00000s (modules-final/config-mcollective)
    00.00000s (modules-final/config-lxd)
    00.00000s (modules-final/config-landscape)
    00.00000s (modules-final/config-chef)
    00.00000s (modules-config/config-snap_config)
    00.00000s (modules-config/config-ntp)
    00.00000s (modules-config/config-emit_upstart)
    00.00000s (modules-config/config-disable-ec2-metadata)
    00.00000s (init-network/setup-datasource)

1 boot records analyzed

Show

The show action is similar to systemd-analyze critical-chain which prints a list of units, the time they started and how long they took. Cloud-init has four stages and within each stage a number of modules may run depending on configuration. cloudinit-analyze show will, for each boot, print this information and a summary total time, per boot.

The following is an abbreviated example of the show output:

$ cloud-init analyze show
-- Boot Record 01 --
The total time elapsed since completing an event is printed after the "@" character.
The time the event takes is printed after the "+" character.

Starting stage: init-local
|``->no cache found @00.01700s +00.00200s
|`->found local data from DataSourceNoCloud @00.11000s +00.40300s
Finished stage: (init-local) 00.94200 seconds

Starting stage: init-network
|`->restored from cache with run check: DataSourceNoCloud [seed=/dev/sr0][dsmode=net] @04.79500s +00.02100s
|`->setting up datasource @04.88900s +00.00000s
|`->reading and applying user-data @04.90100s +00.00400s
|`->reading and applying vendor-data @04.90500s +00.00800s
|`->activating datasource @04.95200s +00.00400s
Finished stage: (init-network) 02.72100 seconds

Starting stage: modules-config
|`->config-emit_upstart ran successfully @15.43100s +00.00000s
|`->config-snap ran successfully @15.43100s +00.00100s
...
|`->config-runcmd ran successfully @16.22300s +00.00100s
|`->config-byobu ran successfully @16.23400s +00.00100s
Finished stage: (modules-config) 00.83500 seconds

Starting stage: modules-final
|`->config-snappy ran successfully @16.87400s +00.00200s
|`->config-package-update-upgrade-install ran successfully @16.87600s +00.00100s
...
|`->config-final-message ran successfully @16.93700s +00.00500s
|`->config-power-state-change ran successfully @16.94300s +00.00000s
Finished stage: (modules-final) 00.10300 seconds

Total Time: 4.60100 seconds

1 boot records analyzed

If additional boot records are detected then they are printed out from oldest to newest.

Dump

The dump action simply dumps the cloud-init logs that the analyze module is performing the analysis on and returns a list of dictionaries that can be consumed for other reporting needs. Each element in the list is a boot entry.

$ cloud-init analyze dump
[
{
  "description": "starting search for local datasources",
  "event_type": "start",
  "name": "init-local",
  "origin": "cloudinit",
  "timestamp": 1567057578.037
},
{
  "description": "attempting to read from cache [check]",
  "event_type": "start",
  "name": "init-local/check-cache",
  "origin": "cloudinit",
  "timestamp": 1567057578.054
},
{
  "description": "no cache found",
  "event_type": "finish",
  "name": "init-local/check-cache",
  "origin": "cloudinit",
  "result": "SUCCESS",
  "timestamp": 1567057578.056
},
{
  "description": "searching for local data from DataSourceNoCloud",
  "event_type": "start",
  "name": "init-local/search-NoCloud",
  "origin": "cloudinit",
  "timestamp": 1567057578.147
},
{
  "description": "found local data from DataSourceNoCloud",
  "event_type": "finish",
  "name": "init-local/search-NoCloud",
  "origin": "cloudinit",
  "result": "SUCCESS",
  "timestamp": 1567057578.55
},
{
  "description": "searching for local datasources",
  "event_type": "finish",
  "name": "init-local",
  "origin": "cloudinit",
  "result": "SUCCESS",
  "timestamp": 1567057578.979
},
{
  "description": "searching for network datasources",
  "event_type": "start",
  "name": "init-network",
  "origin": "cloudinit",
  "timestamp": 1567057582.814
},
{
  "description": "attempting to read from cache [trust]",
  "event_type": "start",
  "name": "init-network/check-cache",
  "origin": "cloudinit",
  "timestamp": 1567057582.832
},
...
{
  "description": "config-power-state-change ran successfully",
  "event_type": "finish",
  "name": "modules-final/config-power-state-change",
  "origin": "cloudinit",
  "result": "SUCCESS",
  "timestamp": 1567057594.98
},
{
  "description": "running modules for final",
  "event_type": "finish",
  "name": "modules-final",
  "origin": "cloudinit",
  "result": "SUCCESS",
  "timestamp": 1567057594.982
}
]

Boot

The boot action prints out kernel related timestamps that are not included in any of the cloud-init logs. There are three different timestamps that are presented to the user:

• kernel start

• kernel finish boot

• cloud-init start

This was added for additional clarity into the boot process that cloud-init does not have control over, to aid in debugging of performance issues related to cloud-init startup, and tracking regression.

$ cloud-init analyze boot
-- Most Recent Boot Record --
    Kernel Started at: 2019-08-29 01:35:37.753790
    Kernel ended boot at: 2019-08-29 01:35:38.807407
    Kernel time to boot (seconds): 1.053617000579834
    Cloud-init activated by systemd at: 2019-08-29 01:35:43.992460
    Time between Kernel end boot and Cloud-init activation (seconds): 5.185053110122681
    Cloud-init start: 2019-08-29 08:35:45.867000
successful

Timestamp Gathering

The following boot related timestamps are gathered on demand when cloud-init analyze boot runs:

• Kernel startup gathered from system uptime

• Kernel finishes initialization from systemd UserSpaceMonotonicTimestamp property

• Cloud-init activation from the property InactiveExitTimestamp of the cloud-init local systemd unit

In order to gather the necessary timestamps using systemd, running the commands below will gather the UserspaceTimestamp and InactiveExitTimestamp:

$ systemctl show -p UserspaceTimestampMonotonic
UserspaceTimestampMonotonic=989279
$ systemctl show cloud-init-local -p InactiveExitTimestampMonotonic
InactiveExitTimestampMonotonic=4493126

The UserspaceTimestamp tracks when the init system starts, which is used as an indicator of kernel finishing initialization. The InactiveExitTimestamp tracks when a particular systemd unit transitions from the Inactive to Active state, which can be used to mark the beginning of systemd’s activation of cloud-init.

Currently this only works for distros that use systemd as the init process. We will be expanding support for other distros in the future and this document will be updated accordingly.

If systemd is not present on the system, dmesg is used to attempt to find an event that logs the beginning of the init system. However, with this method only the first two timestamps are able to be found; dmesg does not monitor userspace processes, so no cloud-init start timestamps are emitted like when using systemd.

Docs

These docs are hosted on Read the Docs. The following will explain how to contribute to and build these docs locally.

The documentation is primarily written in reStructuredText.

Building

There is a makefile target to build the documentation for you:

$ tox -e doc

This will do two things:

• Build the documentation using sphinx

• Run doc8 against the documentation source code

Once build the HTML files will be viewable in doc/rtd_html. Use your web browser to open index.html to view and navigate the site.

Style Guide

Headings

The headings used across the documentation use the following hierarchy:

• *****: used once atop of a new page

• =====: each sections on the page

• -----: subsections

• ^^^^^: sub-subsections

• """"": paragraphs

The top level header ###### is reserved for the first page.

If under and overline are used, their length must be identical. The length of the underline must be at least as long as the title itself

Line Length

Please keep the line lengths to a maximum of 79 characters. This ensures that the pages and tables do not get too wide that side scrolling is required.

Header

Adding a link at the top of the page allows for the page to be referenced by other pages. For example for the FAQ page this would be:

.. _faq:

Vertical Whitespace

One newline between each section helps ensure readability of the documentation source code.

Common Words

There are some common words that should follow specific usage:

• cloud-init: always lower case with a hyphen, unless starting a sentence in which case only the ‘C’ is capitalized (e.g. Cloud-init).

• metadata: one word

• user data: two words, not to be combined

• vendor data: like user data, it is two words

Testing

cloud-init has both unit tests and integration tests. Unit tests can be found at tests/unittests. Integration tests can be found at tests/integration_tests. Documentation specifically for integration tests can be found on the Integration Testing page, but the guidelines specified below apply to both types of tests.

cloud-init uses pytest to run its tests, and has tests written both as unittest.TestCase sub-classes and as un-subclassed pytest tests.

Guidelines

The following guidelines should be followed.

Test Layout

• For ease of organisation and greater accessibility for developers not familiar with pytest, all cloud-init unit tests must be contained within test classes

  • Put another way, module-level test functions should not be used

• As all tests are contained within classes, it is acceptable to mix TestCase test classes and pytest test classes within the same test file

  • These can be easily distinguished by their definition: pytest classes will not use inheritance at all (e.g. TestGetPackageMirrorInfo), whereas TestCase classes will subclass (indirectly) from TestCase (e.g. TestPrependBaseCommands)

• Unit tests and integration tests are located under cloud-init/tests

  • For consistency, unit test files should have a matching name and directory location under tests/unittests

  • For example: the expected test file for code in cloudinit/path/to/file.py is tests/unittests/path/to/test_file.py

pytest Tests

• pytest test classes should use pytest fixtures to share functionality instead of inheritance

• pytest tests should use bare assert statements, to take advantage of pytest’s assertion introspection

pytest Version Gotchas

As we still support Ubuntu 18.04 (Bionic Beaver), we can only use pytest features that are available in v3.3.2. This is an inexhaustive list of ways in which this may catch you out:

• Only the following built-in fixtures are available 1:

  • cache

  • capfd

  • capfdbinary

  • caplog

  • capsys

  • capsysbinary

  • doctest_namespace

  • monkeypatch

  • pytestconfig

  • record_xml_property

  • recwarn

  • tmpdir_factory

  • tmpdir

Mocking and Assertions

• Variables/parameter names for Mock or MagicMock instances should start with m_ to clearly distinguish them from non-mock variables

  • For example, m_readurl (which would be a mock for readurl)

• The assert_* methods that are available on Mock and MagicMock objects should be avoided, as typos in these method names may not raise AttributeError (and so can cause tests to silently pass). An important exception: if a Mock is autospecced then misspelled assertion methods will raise an AttributeError, so these assertion methods may be used on autospecced Mock objects.

  For non-autospecced Mock s, these substitutions can be used (m is assumed to be a Mock):

  • m.assert_any_call(*args, **kwargs) => assert mock.call(*args, **kwargs) in m.call_args_list

  • m.assert_called() => assert 0 != m.call_count

  • m.assert_called_once() => assert 1 == m.call_count

  • m.assert_called_once_with(*args, **kwargs) => assert [mock.call(*args, **kwargs)] == m.call_args_list

  • m.assert_called_with(*args, **kwargs) => assert mock.call(*args, **kwargs) == m.call_args_list[-1]

  • m.assert_has_calls(call_list, any_order=True) => for call in call_list: assert call in m.call_args_list

    • m.assert_has_calls(...) and m.assert_has_calls(..., any_order=False) are not easily replicated in a single statement, so their use when appropriate is acceptable.

  • m.assert_not_called() => assert 0 == m.call_count

• When there are multiple patch calls in a test file for the module it is testing, it may be desirable to capture the shared string prefix for these patch calls in a module-level variable. If used, such variables should be named M_PATH or, for datasource tests, DS_PATH.

Test Argument Ordering

• Test arguments should be ordered as follows:

  • mock.patch arguments. When used as a decorator, mock.patch partially applies its generated Mock object as the first argument, so these arguments must go first.

  • pytest.mark.parametrize arguments, in the order specified to the parametrize decorator. These arguments are also provided by a decorator, so it’s natural that they sit next to the mock.patch arguments.

  • Fixture arguments, alphabetically. These are not provided by a decorator, so they are last, and their order has no defined meaning, so we default to alphabetical.

• It follows from this ordering of test arguments (so that we retain the property that arguments left-to-right correspond to decorators bottom-to-top) that test decorators should be ordered as follows:

  • pytest.mark.parametrize

  • mock.patch

1

This list of fixtures (with markup) can be reproduced by running:

python3 -m pytest  --fixtures -q | grep "^[^ -]" | grep -v 'no tests ran in' | sort | sed 's/ \[session scope\]//g;s/.*/* ``\0``/g'

in an ubuntu lxd container with python3-pytest installed.

Integration Testing

Overview

Integration tests are written using pytest and are located at tests/integration_tests. General design principles laid out in Testing should be followed for integration tests.

Setup is accomplished via a set of fixtures located in tests/integration_tests/conftest.py.

Test Definition

Tests are defined like any other pytest test. The user_data mark can be used to supply the cloud-config user data. Platform specific marks can be used to limit tests to particular platforms. The client fixture can be used to interact with the launched test instance.

See Examples section for examples.

Test Execution

Test execution happens via pytest. A tox definition exists to run integration tests. To run all integration tests, you would run:

$ tox -e integration-tests

Pytest arguments may also be passed. For example:

$ tox -e integration-tests tests/integration_tests/modules/test_combined.py

Configuration

All possible configuration values are defined in tests/integration_tests/integration_settings.py. Defaults can be overridden by supplying values in tests/integration_tests/user_settings.py or by providing an environment variable of the same name prepended with CLOUD_INIT_. For example, to set the PLATFORM setting:

CLOUD_INIT_PLATFORM='ec2' pytest tests/integration_tests/

Cloud Interation

Cloud interaction happens via the pycloudlib library. In order to run integration tests, pycloudlib must first be configured.

For a minimal setup using LXD, write the following to ~/.config/pycloudlib.toml:

[lxd]

Image Selection

Each integration testing run uses a single image as its basis. This image is configured using the OS_IMAGE variable; see Configuration for details of how configuration works.

OS_IMAGE can take two types of value: an Ubuntu series name (e.g. “focal”), or an image specification. If an Ubuntu series name is given, then the most recent image for that series on the target cloud will be used. For other use cases, an image specification is used.

In its simplest form, an image specification can simply be a cloud’s image ID (e.g. “ami-deadbeef”, “ubuntu:focal”). In this case, the image so-identified will be used as the basis for this testing run.

This has a drawback, however: as we do not know what OS or release is within the image, the integration testing framework will run all tests against the image in question. If it’s a RHEL8 image, then we would expect Ubuntu-specific tests to fail (and vice versa).

To address this, a full image specification can be given. This is of the form: <image_id>[::<os>[::<release>]] where image_id is a cloud’s image ID, os is the OS name, and release is the OS release name. So, for example, Ubuntu 18.04 (Bionic Beaver) on LXD is ubuntu:bionic::ubuntu::bionic or RHEL 8 on Amazon is ami-justanexample::rhel::8. When a full specification is given, only tests which are intended for use on that OS and release will be executed.

Image Setup

Image setup occurs once when a test session begins and is implemented via fixture. Image setup roughly follows these steps:

• Launch an instance on the specified test platform

• Install the version of cloud-init under test

• Run cloud-init clean on the instance so subsequent boots resemble out of the box behavior

• Take a snapshot of the instance to be used as a new image from which new instances can be launched

Test Setup

Test setup occurs between image setup and test execution. Test setup is implemented via one of the client fixtures. When a client fixture is used, a test instance from which to run tests is launched prior to test execution and torn down after.

Continuous Integration

A subset of the integration tests are run when a pull request is submitted on Github. The tests run on these continuous integration (CI) runs are given a pytest mark:

@pytest.mark.ci

Most new tests should not use this mark, so be aware that having a successful CI run does not necessarily mean that your test passed successfully.

Fixtures

Integration tests rely heavily on fixtures to do initial test setup. One or more of these fixtures will be used in almost every integration test.

Details such as the cloud platform or initial image to use are determined via what is specified in the Configuration.

client

The client fixture should be used for most test cases. It ensures:

• All setup performed by session_cloud and setup_image

• Pytest marks used during instance creation are obtained and applied

• The test instance is launched

• Test failure status is determined after test execution

• Logs are collected (if configured) after test execution

• The test instance is torn down after test execution

module_client and class_client fixtures also exist for the purpose of running multiple tests against a single launched instance. They provide the exact same functionality as client, but are scoped to the module or class respectively.

session_cloud

The session_cloud session-scoped fixture will provide an IntegrationCloud instance for the currently configured cloud. The fixture also ensures that any custom cloud session cleanup is performed.

setup_image

The setup_image session-scope fixture will create a new image to launch all further cloud instances during this test run. It ensures:

• A cloud instance is launched on the configured platform

• The version of cloud-init under test is installed on the instance

• cloud-init clean --logs is run on the instance

• A snapshot of the instance is taken to be used as the basis for future instance launches

• The originally launched instance is torn down

• The custom created image is torn down after all tests finish

Examples

A simple test case using the client fixture:

USER_DATA = """\
#cloud-config
bootcmd:
- echo 'hello!' > /var/tmp/hello.txt
"""


@pytest.mark.user_data(USER_DATA)
def test_bootcmd(client):
    log = client.read_from_file("/var/log/cloud-init.log")
    assert "Shellified 1 commands." in log
    assert client.execute('cat /var/tmp/hello.txt').strip() == "hello!"

Customizing the launch arguments before launching an instance manually:

def test_launch(session_cloud: IntegrationCloud, setup_image):
    with session_cloud.launch(launch_kwargs={"wait": False}) as client:
        client.instance.wait()
        assert client.execute("echo hello world").strip() == "hello world"