Additionally, *group_names* is a list (array) of all the groups the current host is in. This can be used in templates using Jinja2 syntax to make template source files that vary based on the group membership (or role) of the host::
{% if 'webserver' in group_names %}
# some part of a configuration file that only applies to webservers
{% endif %}
*groups* is a list of all the groups (and hosts) in the inventory. This can be used to enumerate all hosts within a group.
For example::
{% for host in groups['app_servers'] %}
# something that applies to all app servers.
{% endfor %}
A frequently used idiom is walking a group to find all IP addresses in that group::
An example of this could include pointing a frontend proxy server to all of the app servers, setting up the correct firewall rules between servers, etc.
Just a few other 'magic' variables are available... There aren't many.
Additionally, *inventory_hostname* is the name of the hostname as configured in Ansible's inventory host file. This can
be useful for when you don't want to rely on the discovered hostname `ansible_hostname` or for other mysterious
reasons. If you have a long FQDN, *inventory_hostname_short* also contains the part up to the first
period, without the rest of the domain.
Don't worry about any of this unless you think you need it. You'll know when you do.
Also available, *inventory_dir* is the pathname of the directory holding Ansible's inventory host file, *inventory_file* is the pathname and the filename pointing to the Ansible's inventory host file.
Variable File Separation
````````````````````````
It's a great idea to keep your playbooks under source control, but
you may wish to make the playbook source public while keeping certain
important variables private. Similarly, sometimes you may just
want to keep certain information in different files, away from
the main playbook.
You can do this by using an external variables file, or files, just like this::
---
- hosts: all
remote_user: root
vars:
favcolor: blue
vars_files:
- /vars/external_vars.yml
tasks:
- name: this is just a placeholder
command: /bin/echo foo
This removes the risk of sharing sensitive data with others when
sharing your playbook source with them.
The contents of each variables file is a simple YAML dictionary, like this::
---
# in the above example, this would be vars/external_vars.yml
somevar: somevalue
password: magic
.. note::
It's also possible to keep per-host and per-group variables in very
similar files, this is covered in :ref:`patterns`.
Prompting For Sensitive Data
````````````````````````````
You may wish to prompt the user for certain input, and can
do so with the similarly named 'vars_prompt' section. This has uses
beyond security, for instance, you may use the same playbook for all
software releases and would prompt for a particular release version
in a push-script::
---
- hosts: all
remote_user: root
vars:
from: "camelot"
vars_prompt:
name: "what is your name?"
quest: "what is your quest?"
favcolor: "what is your favorite color?"
There are full examples of both of these items in the github examples/playbooks directory.
If you have a variable that changes infrequently, it might make sense to
provide a default value that can be overridden. This can be accomplished using
the default argument::
vars_prompt:
- name: "release_version"
prompt: "Product release version"
default: "1.0"
An alternative form of vars_prompt allows for hiding input from the user, and may later support
some other options, but otherwise works equivalently::
vars_prompt:
- name: "some_password"
prompt: "Enter password"
private: yes
- name: "release_version"
prompt: "Product release version"
private: no
If `Passlib <http://pythonhosted.org/passlib/>`_ is installed, vars_prompt can also crypt the
entered value so you can use it, for instance, with the user module to define a password::
vars_prompt:
- name: "my_password2"
prompt: "Enter password2"
private: yes
encrypt: "md5_crypt"
confirm: yes
salt_size: 7
You can use any crypt scheme supported by 'Passlib':
As with the case of 'with_items' above, you can use previously defined variables. Just specify the variable's name without templating it with '{{ }}'::
- name: here, 'users' contains the above list of employees
``with_password`` and associated lookup macro generate a random plaintext password and store it in
a file at a given filepath. Support for crypted save modes (as with vars_prompt) is pending. If the
file exists previously, it will retrieve its contents, behaving just like with_file. Usage of variables like "{{ inventory_hostname }}" in the filepath can be used to set
up random passwords per host (what simplifies password management in 'host_vars' variables).
Generated passwords contain a random mix of upper and lowercase ASCII letters, the
numbers 0-9 and punctuation (". , : - _"). The default length of a generated password is 30 characters.
This length can be changed by passing an extra parameter::
It is quite possible that you may need to get package updates through a proxy, or even get some package
updates through a proxy and access other packages not through a proxy. Ansible makes it easy for you
to configure your environment by using the 'environment' keyword. Here is an example::
- hosts: all
remote_user: root
tasks:
- apt: name=cobbler state=installed
environment:
http_proxy: http://proxy.example.com:8080
The environment can also be stored in a variable, and accessed like so::
- hosts: all
remote_user: root
# here we make a variable named "env" that is a dictionary
vars:
proxy_env:
http_proxy: http://proxy.example.com:8080
tasks:
- apt: name=cobbler state=installed
environment: "{{ proxy_env }}"
While just proxy settings were shown above, any number of settings can be supplied. The most logical place
to define an environment hash might be a group_vars file, like so::
---
# file: group_vars/boston
ntp_server: ntp.bos.example.com
backup: bak.bos.example.com
proxy_env:
http_proxy: http://proxy.bos.example.com:8080
https_proxy: http://proxy.bos.example.com:8080
Getting values from files
`````````````````````````
.. versionadded:: 0.8
Sometimes you'll want to include the content of a file directly into a playbook. You can do so using a macro.
This syntax will remain in future versions, though we will also will provide ways to do this via lookup plugins (see "More Loops") as well. What follows
is an example using the authorized_key module, which requires the actual text of the SSH key as a parameter::
Sometimes a configuration file you want to copy, or a template you will use may depend on a variable.
The following construct selects the first available file appropriate for the variables of a given host, which is often much cleaner than putting a lot of if conditionals in a template.
The following example shows how to template out a configuration file that was very different between, say, CentOS and Debian::
first_available_file is only available to the copy and template modules.
Asynchronous Actions and Polling
````````````````````````````````
By default tasks in playbooks block, meaning the connections stay open
until the task is done on each node. If executing playbooks with
a small parallelism value (aka ``--forks``), you may wish that long
running operations can go faster. The easiest way to do this is
to kick them off all at once and then poll until they are done.
You will also want to use asynchronous mode on very long running
operations that might be subject to timeout.
To launch a task asynchronously, specify its maximum runtime
and how frequently you would like to poll for status. The default
poll value is 10 seconds if you do not specify a value for `poll`::
---
- hosts: all
remote_user: root
tasks:
- name: simulate long running op (15 sec), wait for up to 45, poll every 5
command: /bin/sleep 15
async: 45
poll: 5
.. note::
There is no default for the async time limit. If you leave off the
'async' keyword, the task runs synchronously, which is Ansible's
default.
Alternatively, if you do not need to wait on the task to complete, you may
"fire and forget" by specifying a poll value of 0::
---
- hosts: all
remote_user: root
tasks:
- name: simulate long running op, allow to run for 45, fire and forget
command: /bin/sleep 15
async: 45
poll: 0
.. note::
You shouldn't "fire and forget" with operations that require
exclusive locks, such as yum transactions, if you expect to run other
commands later in the playbook against those same resources.
.. note::
Using a higher value for ``--forks`` will result in kicking off asynchronous
tasks even faster. This also increases the efficiency of polling.
Local Playbooks
```````````````
It may be useful to use a playbook locally, rather than by connecting over SSH. This can be useful
for assuring the configuration of a system by putting a playbook on a crontab. This may also be used
to run a playbook inside a OS installer, such as an Anaconda kickstart.
To run an entire playbook locally, just set the "hosts:" line to "hosts:127.0.0.1" and then run the playbook like so::
ansible-playbook playbook.yml --connection=local
Alternatively, a local connection can be used in a single playbook play, even if other plays in the playbook
use the default remote connection type::
hosts: 127.0.0.1
connection: local
Turning Off Facts
`````````````````
If you know you don't need any fact data about your hosts, and know everything about your systems centrally, you
can turn off fact gathering. This has advantages in scaling Ansible in push mode with very large numbers of
systems, mainly, or if you are using Ansible on experimental platforms. In any play, just do this::
- hosts: whatever
gather_facts: no
Pull-Mode Playbooks
```````````````````
The use of playbooks in local mode (above) is made extremely powerful with the addition of `ansible-pull`.
A script for setting up ansible-pull is provided in the examples/playbooks directory of the source
checkout.
The basic idea is to use Ansible to set up a remote copy of Ansible on each managed node, each set to run via
cron and update playbook source via git. This inverts the default push architecture of Ansible into a pull
architecture, which has near-limitless scaling potential. The setup playbook can be tuned to change
the cron frequency, logging locations, and parameters to ansible-pull.
This is useful both for extreme scale-out as well as periodic remediation. Usage of the 'fetch' module to retrieve
logs from ansible-pull runs would be an excellent way to gather and analyze remote logs from ansible-pull.
Register Variables
``````````````````
.. versionadded:: 0.7
Often in a playbook it may be useful to store the result of a given command in a variable and access
it later. Use of the command module in this way can in many ways eliminate the need to write site specific facts, for
instance, you could test for the existence of a particular program.
The 'register' keyword decides what variable to save a result in. The resulting variables can be used in templates, action lines, or *when* statements. It looks like this (in an obviously trivial example)::
- name: test play
hosts: all
tasks:
- shell: cat /etc/motd
register: motd_contents
- shell: echo "motd contains the word hi"
when: motd_contents.stdout.find('hi') != -1
As shown previously, the registered variable's string contents are accessible with the 'stdout' value.
The registered result can be used in the "with_items" of a task if it is converted into
a list (or already is a list) as shown below. "stdout_lines" is already available on the object as
well though you could also call "home_dirs.stdout.split()" if you wanted, and could split by other
fields::
- name: registered variable usage as a with_items list
A common pattern is to use a local action to call 'rsync' to recursively copy files to the managed servers.
Here is an example::
---
# ...
tasks:
- name: recursively copy files from management server to target
local_action: command rsync -a /path/to/files {{ inventory_hostname }}:/path/to/target/
Note that you must have passphrase-less SSH keys or an ssh-agent configured for this to work, otherwise rsync
will need to ask for a passphrase.
Accelerated Mode
````````````````
.. versionadded:: 1.3
While SSH using the ControlPersist feature is quite fast and scalable, there is a certain amount of overhead involved in
creating connections. This can become something of a bottleneck when the number of hosts grows into the hundreds or
thousands. To help overcome this, Ansible offers an accelerated connection option. Accelerated mode can be anywhere from
2-6x faster than SSH with ControlPersist enabled, and 10x faster than paramiko.
Accelerated mode works by launching a temporary daemon over SSH. Once the daemon is running, Ansible will connect directly
to it via a raw socket connection. Ansible secures this communication by using a temporary AES key that is uploaded during
the SSH connection (this key is different for every host, and is also regenerated every time the daemon is started). By default,
Ansible will use port 5099 for the accelerated connection, though this is configurable. Once running, the daemon will accept
connections for 30 minutes, after which time it will terminate itself and need to be restarted over SSH.
Accelerated mode offers several improvments over the original fireball mode:
* No bootstrapping is required, only a single line needs to be added to each play you wish to run in accelerated mode.
* Support for sudo commands (see below for more details and caveats).
* Fewer requirements! ZeroMQ is no longer required, nor are there any special packages beyond python-keyczar.
In order to use accelerated mode, simply add `accelerate: true` to your play::
---
- hosts: all
accelerate: true
tasks:
- name: some task
command: echo {{ item }}
with_items:
- foo
- bar
- baz
If you wish to change the port Ansible will use for the accelerated connection, just add the `accelerated_port` option::
---
- hosts: all
accelerate: true
# default port is 5099
accelerate_port: 10000
The `accelerate_port` option can also be specified in the environment variable ACCELERATE_PORT, or in your `ansible.cfg` configuration::
[accelerate]
accelerate_port = 5099
As noted above, accelerated mode also supports running tasks via sudo, however there are two important caveats:
* You must remove requiretty from your sudoers options.
* Prompting for the sudo password is not yet supported, so the NOPASSWD option is required for commands.
Fireball Mode
`````````````
.. versionadded:: 0.8 (deprecated as of 1.3)
.. note::
The following section has been deprecated as of Ansible 1.3 in favor of the accelerated mode described above. This
documentation is here for users who may still be using the original fireball connection method only, and should not
be used for any new deployments.
Ansible's core connection types of 'local', 'paramiko', and 'ssh' are augmented in version 0.8 and later by a new extra-fast
connection type called 'fireball'. It can only be used with playbooks and does require some additional setup
outside the lines of Ansible's normal "no bootstrapping" philosophy. You are not required to use fireball mode
to use Ansible, though some users may appreciate it.
Fireball mode works by launching a temporary 0mq daemon from SSH that by default lives for only 30 minutes before
shutting off. Fireball mode, once running, uses temporary AES keys to encrypt a session, and requires direct
communication to given nodes on the configured port. The default is 5099. The fireball daemon runs as any user you
set it down as. So it can run as you, root, or so on. If multiple users are running Ansible as the same batch of hosts,
take care to use unique ports.
Fireball mode is roughly 10 times faster than paramiko for communicating with nodes and may be a good option
if you have a large number of hosts::
---
# set up the fireball transport
- hosts: all
gather_facts: no
connection: ssh # or paramiko
sudo: yes
tasks:
- action: fireball
# these operations will occur over the fireball transport
- hosts: all
connection: fireball
tasks:
- shell: echo "Hello {{ item }}"
with_items:
- one
- two
In order to use fireball mode, certain dependencies must be installed on both ends. You can use this playbook as a basis for initial bootstrapping on
any platform. You will also need gcc and zeromq-devel installed from your package manager, which you can of course also get Ansible to install::
---
- hosts: all
sudo: yes
gather_facts: no
connection: ssh
tasks:
- easy_install: name=pip
- pip: name={{ item }} state=present
with_items:
- pyzmq
- pyasn1
- PyCrypto
- python-keyczar
Fedora and EPEL also have Ansible RPM subpackages available for fireball-dependencies.
Also see the module documentation section.
Understanding Variable Precedence
`````````````````````````````````
You have already learned about inventory variables, 'vars', and 'vars_files'. In the
event the same variable name occurs in more than one place, what happens? There are really three tiers
of precedence, and within those tiers, some minor ordering rules that you probably won't even need to remember.
We'll explain them anyway though.
Variables that are set during the execution of the play have highest priority. This includes registered
variables and facts, which are discovered pieces of information about remote hosts.
Descending in priority are variables defined in the playbook. 'vars_files' as defined in the playbook are next up,
followed by variables as passed to ansible-playbook via --extra-vars (-e), then variables defined in the 'vars' section. These
should all be taken to be basically the same thing -- good places to define constants about what the play does to all hosts
in the play.
Finally, inventory variables have the least priority. Variables about hosts override those about groups.
If a variable is defined in multiple groups and one group is a child of the other, the child group variable
will override the variable set in the parent.
This makes the 'group_vars/all' file the best place to define a default value you wish to override in another
group, or even in a playbook. For example, your organization might set a default ntp server in group_vars/all
and then override it based on a group based on a geographic region. However if you type 'ntpserver: asdf.example.com'
in a vars section of a playbook, you know from reading the playbook that THAT specific value is definitely the one
that is going to be used. You won't be fooled by some variable from inventory sneaking up on you.
So, in short, if you want something easy to remember: facts beat playbook definitions, and
playbook definitions beat inventory variables.
There's a little bit more if you are using roles -- roles fit in the "playbook definitions" category of scale. They are
trumped by facts, and still trump inventory variables. However, there's a bit of extra magic.
Variables passed as parameters to the role are accesible only within that role (and dependencies of that role). You can
almost think of them like programming functions or macros.
Variables loaded via the 'vars/' directory of a role are made available to all roles and tasks, which in older versions of Ansible
could be confusing in the case of a reused variable name. In Ansible 1.3 and later, however, vars/ directories are guaranteed to be scoped to the current role, just like roles parameters. They are still available globally though, so if you want to set a variable like "ntp_server" in a common role, other roles can still make use of it. Thus they are just like "vars_files" construct that they emulate, but they have a bit more of a "Do What I Mean" semantic to them. They are smarter.
If there are role dependencies involved, dependent roles can set variables visible to the roles that require them, but
the requiring role is allowed to override those variables. For instance if a role "myapp" requires "apache", and
the value of "apache_port" in "apache" is 80, "myapp" could choose to set it to 8080. Thus you may think of this somewhat
like an inheritance system if you're a developer -- though it's not exactly -- and we don't require folks to think in programming terms to know how things work.
If you want, you can choose to prefix variable names with the name of your role and be extra sure of where
data sources are coming from, but this is optional. However it can be a nice thing to do in your templates as you immediately
know where the variable was defined.
Ultimately, the variable system may seem complex -- but it's really not. It's mostly a "Do What I Mean" kind of system, though knowing the details may help you if you get stuck or are trying to do something advanced. Feel free to experiment!
Check Mode ("Dry Run") --check
```````````````````````````````
.. versionadded:: 1.1
When ansible-playbook is executed with --check it will not make any changes on remote systems. Instead, any module
instrumented to support 'check mode' (which contains the primary core modules, but it is not required that all modules do
this) will report what changes they would have made. Other modules that do not support check mode will also take no
action, but just will not report what changes they might have made.
Check mode is just a simulation, and if you have steps that use conditionals that depend on the results of prior commands,
it may be less useful for you. However it is great for one-node-at-time basic configuration management use cases.
Example::
ansible-playbook foo.yml --check
Running a task in check mode
````````````````````````````
.. versionadded:: 1.3
Sometimes you may want to have a task to be executed even in check
mode. To achieve this, use the `always_run` clause on the task. Its
value is a Jinja2 expression, just like the `when` clause. In simple
cases a boolean YAML value would be sufficient as a value.
Example::
tasks:
- name: this task is run even in check mode
command: /something/to/run --even-in-check-mode
always_run: yes
As a reminder, a task with a `when` clause evaluated to false, will
still be skipped even if it has a `always_run` clause evaluated to
true.
Showing Differences with --diff
```````````````````````````````
.. versionadded:: 1.1
The --diff option to ansible-playbook works great with --check (detailed above) but can also be used by itself. When this flag is supplied, if any templated files on the remote system are changed, and the ansible-playbook CLI will report back
the textual changes made to the file (or, if used with --check, the changes that would have been made). Since the diff
feature produces a large amount of output, it is best used when checking a single host at a time, like so::
As discussed in the playbooks chapter, Ansible facts are a way of getting data about remote systems for use in playbook variables.
Usually these are discovered automatically by the 'setup' module in Ansible. Users can also write custom facts modules, as described
in the API guide. However, what if you want to have a simple way to provide system or user
provided data for use in Ansible variables, without writing a fact module? For instance, what if you want users to be able to control some aspect about how their systems are managed? "Facts.d" is one such mechanism.
If a remotely managed system has an "/etc/ansible/facts.d" directory, any files in this directory
ending in ".fact", can be JSON, INI, or executable files returning JSON, and these can supply local facts in Ansible.
For instance assume a /etc/ansible/facts.d/preferences.fact::
[general]
asdf=1
bar=2
This will produce a hash variable fact named "general" with 'asdf' and 'bar' as members.
To validate this, run the following::
ansible <hostname> -m setup -a "filter=ansible_local"
And you will see the following fact added::
"ansible_local": {
"preferences": {
"general": {
"asdf" : "1",
"bar" : "2"
}
}
}
And this data can be accessed in a template/playbook as::
{{ ansible_local.preferences.general.asdf }}
The local namespace prevents any user supplied fact from overriding system facts
or variables defined elsewhere in the playbook.
Style Points
````````````
Ansible playbooks are colorized. If you do not like this, set the ANSIBLE_NOCOLOR=1 environment variable.
Ansible playbooks also look more impressive with cowsay installed, and we encourage installing this package.
.. seealso::
:doc:`YAMLSyntax`
Learn about YAML syntax
:doc:`playbooks`
Review the basic playbook features
:doc:`bestpractices`
Various tips about playbooks in the real world
:doc:`modules`
Learn about available modules
:doc:`moduledev`
Learn how to extend Ansible by writing your own modules
There are several interesting ways to use Ansible from an API perspective. You can use
the Ansible python API to control nodes, you can extend Ansible to respond to various python events,
and you can plug in inventory data from external data sources. Ansible is written in its own
API so you have a considerable amount of power across the board.
API so you have a considerable amount of power across the board. This chapter discusses the Python API.
.. contents:: `Table of contents`
:depth: 2
...
...
@@ -81,317 +81,3 @@ Advanced programmers may also wish to read the source to ansible itself, for
it uses the Runner() API (with all available options) to implement the
command line tools ``ansible`` and ``ansible-playbook``.
Plugins Available Online
------------------------
The remainder of features in the API docs have components available in `ansible-plugins <https://github.com/ansible/ansible/blob/devel/plugins>`_. Send us a github pull request if you develop any interesting features.
External Inventory Scripts
--------------------------
Often a user of a configuration management system will want to keep inventory
in a different system. Frequent examples include LDAP, `Cobbler <http://cobbler.github.com>`_,
or a piece of expensive enterprisey CMDB software. Ansible easily supports all
of these options via an external inventory system. The plugins directory contains some of these already -- including options for EC2/Eucalyptus and OpenStack, which will be detailed below.
It's possible to write an external inventory script in any language. If you are familiar with Puppet terminology, this concept is basically the same as 'external nodes', with the slight difference that it also defines which hosts are managed.
Script Conventions
``````````````````
When the external node script is called with the single argument '--list', the script must return a JSON hash/dictionary of all the groups to be managed.
Each group's value should be either a hash/dictionary containing a list of each host/IP, potential child groups, and potential group variables, or
Before version 1.0, each group could only have a list of hostnames/IP addresses, like the webservers, marietta, and 5points groups above.
When called with the arguments '--host <hostname>' (where <hostname> is a host from above), the script must return either an empty JSON
hash/dictionary, or a hash/dictionary of variables to make available to templates and playbooks. Returning variables is optional,
if the script does not wish to do this, returning an empty hash/dictionary is the way to go::
{
"favcolor" : "red",
"ntpserver" : "wolf.example.com",
"monitoring" : "pack.example.com"
}
Tuning the External Inventory Script
````````````````````````````````````
.. versionadded:: 1.3
The stock inventory script system detailed above works for all versions of Ansible, but calling
'--host' for every host can be rather expensive, especially if it involves expensive API calls to
a remote subsystemm. In Ansible
1.3 or later, if the inventory script returns a top level element called "_meta", it is possible
to return all of the host variables in one inventory script call. When this meta element contains
a value for "hostvars", the inventory script will not be invoked with "--host" for each host. This
results in a significant performance increase for large numbers of hosts, and also makes client
side caching easier to implement for the inventory script.
The data to be added to the top level JSON dictionary looks like this::
{
# results of inventory script as above go here
# ...
"_meta" : {
"hostvars" : {
"moocow.example.com" : { "asdf" : 1234 },
"llama.example.com" : { "asdf" : 5678 },
}
}
}
Example: The Cobbler External Inventory Script
``````````````````````````````````````````````
It is expected that many Ansible users will also be `Cobbler <http://cobbler.github.com>`_ users. Cobbler has a generic
layer that allows it to represent data for multiple configuration management systems (even at the same time), and has
been referred to as a 'lightweight CMDB' by some admins. This particular script will communicate with Cobbler
using Cobbler's XMLRPC API.
To tie Ansible's inventory to Cobbler (optional), copy `this script <https://raw.github.com/ansible/ansible/devel/plugins/inventory/cobbler.py>`_ to /etc/ansible/hosts and `chmod +x` the file. cobblerd will now need
to be running when you are using Ansible.
Test the file by running `./etc/ansible/hosts` directly. You should see some JSON data output, but it may not have
anything in it just yet.
Let's explore what this does. In cobbler, assume a scenario somewhat like the following::
cobbler system edit --name=foo --dns-name="foo.example.com" --mgmt-classes="atlanta" --ksmeta="c=4"
cobbler system edit --name=bar --dns-name="bar.example.com" --mgmt-classes="atlanta" --ksmeta="c=5"
In the example above, the system 'foo.example.com' will be addressable by ansible directly, but will also be addressable when using the group names 'webserver' or 'atlanta'. Since Ansible uses SSH, we'll try to contract system foo over 'foo.example.com', only, never just 'foo'. Similarly, if you try "ansible foo" it wouldn't find the system... but "ansible 'foo*'" would, because the system DNS name starts with 'foo'.
The script doesn't just provide host and group info. In addition, as a bonus, when the 'setup' module is run (which happens automatically when using playbooks), the variables 'a', 'b', and 'c' will all be auto-populated in the templates::
# file: /srv/motd.j2
Welcome, I am templated with a value of a={{ a }}, b={{ b }}, and c={{ c }}
Which could be executed just like this::
ansible webserver -m setup
ansible webserver -m template -a "src=/tmp/motd.j2 dest=/etc/motd"
.. note::
The name 'webserver' came from cobbler, as did the variables for
the config file. You can still pass in your own variables like
normal in Ansible, but variables from the external inventory script
will override any that have the same name.
So, with the template above (motd.j2), this would result in the following data being written to /etc/motd for system 'foo'::
Welcome, I am templated with a value of a=2, b=3, and c=4
And on system 'bar' (bar.example.com)::
Welcome, I am templated with a value of a=2, b=3, and c=5
And technically, though there is no major good reason to do it, this also works too::
ansible webserver -m shell -a "echo {{ a }}"
So in other words, you can use those variables in arguments/actions as well. You might use this to name
a conf.d file appropriately or something similar. Who knows?
So that's the Cobbler integration support -- using the cobbler script as an example, it should be trivial to adapt Ansible to pull inventory, as well as variable information, from any data source. If you create anything interesting, please share with the mailing list, and we can keep it in the source code tree for others to use.
Example: AWS EC2 External Inventory Script
``````````````````````````````````````````
If you use Amazon Web Services EC2, maintaining an inventory file might not be the best approach. For this reason, you can use the `EC2 external inventory <https://raw.github.com/ansible/ansible/devel/plugins/inventory/ec2.py>`_ script.
You can use this script in one of two ways. The easiest is to use Ansible's ``-i`` command line option and specify the path to the script.
ansible -i ec2.py -u ubuntu us-east-1d -m ping
The second option is to copy the script to `/etc/ansible/hosts` and `chmod +x` it. You will also need to copy the `ec2.ini <https://raw.github.com/ansible/ansible/devel/plugins/inventory/ec2.ini>`_ file to `/etc/ansible/ec2.ini`. Then you can run ansible as you would normally.
To successfully make an API call to AWS, you will need to configure Boto (the Python interface to AWS). There are a `variety of methods <http://docs.pythonboto.org/en/latest/boto_config_tut.html>`_ available, but the simplest is just to export two environment variables:
export AWS_ACCESS_KEY_ID='AK123'
export AWS_SECRET_ACCESS_KEY='abc123'
You can test the script by itself to make sure your config is correct
cd plugins/inventory
./ec2.py --list
After a few moments, you should see your entire EC2 inventory across all regions in JSON.
Since each region requires its own API call, if you are only using a small set of regions, feel free to edit ``ec2.ini`` and list only the regions you are interested in. There are other config options in ``ec2.ini`` including cache control, and destination variables.
At their heart, inventory files are simply a mapping from some name to a destination address. The default ``ec2.ini`` settings are configured for running Ansible from outside EC2 (from your laptop for example). If you are running Ansible from within EC2, internal DNS names and IP addresses may make more sense than public DNS names. In this case, you can modify the ``destination_variable`` in ``ec2.ini`` to be the private DNS name of an instance. This is particularly important when running Ansible within a private subnet inside a VPC, where the only way to access an instance is via its private IP address. For VPC instances, `vpc_destination_variable` in ``ec2.ini`` provides a means of using which ever `boto.ec2.instance variable <http://docs.pythonboto.org/en/latest/ref/ec2.html#module-boto.ec2.instance>`_ makes the most sense for your use case.
The EC2 external inventory provides mappings to instances from several groups:
Instance ID
These are groups of one since instance IDs are unique.
e.g.
``i-00112233``
``i-a1b1c1d1``
Region
A group of all instances in an AWS region.
e.g.
``us-east-1``
``us-west-2``
Availability Zone
A group of all instances in an availability zone.
e.g.
``us-east-1a``
``us-east-1b``
Security Group
Instances belong to one or more security groups. A group is created for each security group, with all characters except alphanumerics, dashes (-) converted to underscores (_). Each group is prefixed by ``security_group_``
e.g.
``security_group_default``
``security_group_webservers``
``security_group_Pete_s_Fancy_Group``
Tags
Each instance can have a variety of key/value pairs associated with it called Tags. The most common tag key is 'Name', though anything is possible. Each key/value pair is its own group of instances, again with special characters converted to underscores, in the format ``tag_KEY_VALUE``
When the Ansible is interacting with a specific server, the EC2 inventory script is called again with the ``--host HOST`` option. This looks up the HOST in the index cache to get the instance ID, and then makes an API call to AWS to get information about that specific instance. It then makes information about that instance available as variables to your playbooks. Each variable is prefixed by ``ec2_``. Here are some of the variables available:
- ec2_architecture
- ec2_description
- ec2_dns_name
- ec2_id
- ec2_image_id
- ec2_instance_type
- ec2_ip_address
- ec2_kernel
- ec2_key_name
- ec2_launch_time
- ec2_monitored
- ec2_ownerId
- ec2_placement
- ec2_platform
- ec2_previous_state
- ec2_private_dns_name
- ec2_private_ip_address
- ec2_public_dns_name
- ec2_ramdisk
- ec2_region
- ec2_root_device_name
- ec2_root_device_type
- ec2_security_group_ids
- ec2_security_group_names
- ec2_spot_instance_request_id
- ec2_state
- ec2_state_code
- ec2_state_reason
- ec2_status
- ec2_subnet_id
- ec2_tag_Name
- ec2_tenancy
- ec2_virtualization_type
- ec2_vpc_id
Both ``ec2_security_group_ids`` and ``ec2_security_group_names`` are comma-separated lists of all security groups. Each EC2 tag is a variable in the format ``ec2_tag_KEY``.
To see the complete list of variables available for an instance, run the script by itself::
Though not detailed here in as much depth as the EC2 module, there's also a OpenStack Compute external inventory source in the plugins directory. It requires the Grizzly release of OpenStack or
later. See the inline comments in the module source for how to use it.
Callback Plugins
----------------
Ansible can be configured via code to respond to external events. This can include enhancing logging, signalling an external software
system, or even (yes, really) making sound effects. Some examples are contained in the plugins directory.
Connection Type Plugins
-----------------------
By default, ansible ships with a 'paramiko' SSH, native ssh (just called 'ssh'), and 'local' connection type, and an accelerated connection type named 'fireball'. All of these can be used
in playbooks and with /usr/bin/ansible to decide how you want to talk to remote machines. The basics of these connection types
are covered in the 'getting started' section. Should you want to extend Ansible to support other transports (SNMP? Message bus?
Carrier Pigeon?) it's as simple as copying the format of one of the existing modules and dropping it into the connection plugins
directory. The value of 'smart' for a connection allows selection of paramiko or openssh based on system capabilities, and chooses
'ssh' if OpenSSH supports ControlPersist, in Ansible 1.2.1 an later. Previous versions did not support 'smart'.
Lookup Plugins
--------------
Language constructs like "with_fileglob" and "with_items" are implemented via lookup plugins. Just like other plugin types, you can write your own.
Vars Plugins
------------
Playbook constructs like 'host_vars' and 'group_vars' work via 'vars' plugins. They inject additional variable
data into ansible runs that did not come from an inventory, playbook, or command line. Note that variables
can also be returned from inventory, so in most cases, you won't need to write or understand vars_plugins.
Filter Plugins
--------------
If you want more Jinja2 filters available in a Jinja2 template (filters like to_yaml and to_json are provided by default), they can be extended by writing a filter plugin.
Distributing Plugins
--------------------
.. versionadded:: 0.8
Plugins are loaded from both Python's site_packages (those that ship with ansible) and a configured plugins directory, which defaults
to /usr/share/ansible/plugins, in a subfolder for each plugin type::
* action_plugins
* lookup_plugins
* callback_plugins
* connection_plugins
* filter_plugins
* vars_plugins
To change this path, edit the ansible configuration file.
In addition, plugins can be shipped in a subdirectory relative to a top-level playbook, in folders named the same as indicated above.
.. seealso::
:doc:`modules`
List of built-in modules
`Mailing List <http://groups.google.com/group/ansible-project>`_
Questions? Help? Ideas? Stop by the list on Google Groups
for (hostname, result) in results['contacted'].items():
if not 'failed' in result:
print "%s >>> %s" % (hostname, result['stdout'])
print "FAILED *******"
for (hostname, result) in results['contacted'].items():
if 'failed' in result:
print "%s >>> %s" % (hostname, result['msg'])
print "DOWN *********"
for (hostname, result) in results['dark'].items():
print "%s >>> %s" % (hostname, result)
Advanced programmers may also wish to read the source to ansible itself, for
it uses the Runner() API (with all available options) to implement the
command line tools ``ansible`` and ``ansible-playbook``.
Plugins Available Online
------------------------
The remainder of features in the API docs have components available in `ansible-plugins <https://github.com/ansible/ansible/blob/devel/plugins>`_. Send us a github pull request if you develop any interesting features.
External Inventory Scripts
--------------------------
Often a user of a configuration management system will want to keep inventory
in a different system. Frequent examples include LDAP, `Cobbler <http://cobbler.github.com>`_,
or a piece of expensive enterprisey CMDB software. Ansible easily supports all
of these options via an external inventory system. The plugins directory contains some of these already -- including options for EC2/Eucalyptus and OpenStack, which will be detailed below.
It's possible to write an external inventory script in any language. If you are familiar with Puppet terminology, this concept is basically the same as 'external nodes', with the slight difference that it also defines which hosts are managed.
Script Conventions
``````````````````
When the external node script is called with the single argument '--list', the script must return a JSON hash/dictionary of all the groups to be managed.
Each group's value should be either a hash/dictionary containing a list of each host/IP, potential child groups, and potential group variables, or
Before version 1.0, each group could only have a list of hostnames/IP addresses, like the webservers, marietta, and 5points groups above.
When called with the arguments '--host <hostname>' (where <hostname> is a host from above), the script must return either an empty JSON
hash/dictionary, or a hash/dictionary of variables to make available to templates and playbooks. Returning variables is optional,
if the script does not wish to do this, returning an empty hash/dictionary is the way to go::
{
"favcolor" : "red",
"ntpserver" : "wolf.example.com",
"monitoring" : "pack.example.com"
}
Tuning the External Inventory Script
````````````````````````````````````
.. versionadded:: 1.3
The stock inventory script system detailed above works for all versions of Ansible, but calling
'--host' for every host can be rather expensive, especially if it involves expensive API calls to
a remote subsystemm. In Ansible
1.3 or later, if the inventory script returns a top level element called "_meta", it is possible
to return all of the host variables in one inventory script call. When this meta element contains
a value for "hostvars", the inventory script will not be invoked with "--host" for each host. This
results in a significant performance increase for large numbers of hosts, and also makes client
side caching easier to implement for the inventory script.
The data to be added to the top level JSON dictionary looks like this::
{
# results of inventory script as above go here
# ...
"_meta" : {
"hostvars" : {
"moocow.example.com" : { "asdf" : 1234 },
"llama.example.com" : { "asdf" : 5678 },
}
}
}
Example: The Cobbler External Inventory Script
``````````````````````````````````````````````
It is expected that many Ansible users will also be `Cobbler <http://cobbler.github.com>`_ users. Cobbler has a generic
layer that allows it to represent data for multiple configuration management systems (even at the same time), and has
been referred to as a 'lightweight CMDB' by some admins. This particular script will communicate with Cobbler
using Cobbler's XMLRPC API.
To tie Ansible's inventory to Cobbler (optional), copy `this script <https://raw.github.com/ansible/ansible/devel/plugins/inventory/cobbler.py>`_ to /etc/ansible/hosts and `chmod +x` the file. cobblerd will now need
to be running when you are using Ansible.
Test the file by running `./etc/ansible/hosts` directly. You should see some JSON data output, but it may not have
anything in it just yet.
Let's explore what this does. In cobbler, assume a scenario somewhat like the following::
cobbler system edit --name=foo --dns-name="foo.example.com" --mgmt-classes="atlanta" --ksmeta="c=4"
cobbler system edit --name=bar --dns-name="bar.example.com" --mgmt-classes="atlanta" --ksmeta="c=5"
In the example above, the system 'foo.example.com' will be addressable by ansible directly, but will also be addressable when using the group names 'webserver' or 'atlanta'. Since Ansible uses SSH, we'll try to contract system foo over 'foo.example.com', only, never just 'foo'. Similarly, if you try "ansible foo" it wouldn't find the system... but "ansible 'foo*'" would, because the system DNS name starts with 'foo'.
The script doesn't just provide host and group info. In addition, as a bonus, when the 'setup' module is run (which happens automatically when using playbooks), the variables 'a', 'b', and 'c' will all be auto-populated in the templates::
# file: /srv/motd.j2
Welcome, I am templated with a value of a={{ a }}, b={{ b }}, and c={{ c }}
Which could be executed just like this::
ansible webserver -m setup
ansible webserver -m template -a "src=/tmp/motd.j2 dest=/etc/motd"
.. note::
The name 'webserver' came from cobbler, as did the variables for
the config file. You can still pass in your own variables like
normal in Ansible, but variables from the external inventory script
will override any that have the same name.
So, with the template above (motd.j2), this would result in the following data being written to /etc/motd for system 'foo'::
Welcome, I am templated with a value of a=2, b=3, and c=4
And on system 'bar' (bar.example.com)::
Welcome, I am templated with a value of a=2, b=3, and c=5
And technically, though there is no major good reason to do it, this also works too::
ansible webserver -m shell -a "echo {{ a }}"
So in other words, you can use those variables in arguments/actions as well. You might use this to name
a conf.d file appropriately or something similar. Who knows?
So that's the Cobbler integration support -- using the cobbler script as an example, it should be trivial to adapt Ansible to pull inventory, as well as variable information, from any data source. If you create anything interesting, please share with the mailing list, and we can keep it in the source code tree for others to use.
Example: AWS EC2 External Inventory Script
``````````````````````````````````````````
If you use Amazon Web Services EC2, maintaining an inventory file might not be the best approach. For this reason, you can use the `EC2 external inventory <https://raw.github.com/ansible/ansible/devel/plugins/inventory/ec2.py>`_ script.
You can use this script in one of two ways. The easiest is to use Ansible's ``-i`` command line option and specify the path to the script.
ansible -i ec2.py -u ubuntu us-east-1d -m ping
The second option is to copy the script to `/etc/ansible/hosts` and `chmod +x` it. You will also need to copy the `ec2.ini <https://raw.github.com/ansible/ansible/devel/plugins/inventory/ec2.ini>`_ file to `/etc/ansible/ec2.ini`. Then you can run ansible as you would normally.
To successfully make an API call to AWS, you will need to configure Boto (the Python interface to AWS). There are a `variety of methods <http://docs.pythonboto.org/en/latest/boto_config_tut.html>`_ available, but the simplest is just to export two environment variables:
export AWS_ACCESS_KEY_ID='AK123'
export AWS_SECRET_ACCESS_KEY='abc123'
You can test the script by itself to make sure your config is correct
cd plugins/inventory
./ec2.py --list
After a few moments, you should see your entire EC2 inventory across all regions in JSON.
Since each region requires its own API call, if you are only using a small set of regions, feel free to edit ``ec2.ini`` and list only the regions you are interested in. There are other config options in ``ec2.ini`` including cache control, and destination variables.
At their heart, inventory files are simply a mapping from some name to a destination address. The default ``ec2.ini`` settings are configured for running Ansible from outside EC2 (from your laptop for example). If you are running Ansible from within EC2, internal DNS names and IP addresses may make more sense than public DNS names. In this case, you can modify the ``destination_variable`` in ``ec2.ini`` to be the private DNS name of an instance. This is particularly important when running Ansible within a private subnet inside a VPC, where the only way to access an instance is via its private IP address. For VPC instances, `vpc_destination_variable` in ``ec2.ini`` provides a means of using which ever `boto.ec2.instance variable <http://docs.pythonboto.org/en/latest/ref/ec2.html#module-boto.ec2.instance>`_ makes the most sense for your use case.
The EC2 external inventory provides mappings to instances from several groups:
Instance ID
These are groups of one since instance IDs are unique.
e.g.
``i-00112233``
``i-a1b1c1d1``
Region
A group of all instances in an AWS region.
e.g.
``us-east-1``
``us-west-2``
Availability Zone
A group of all instances in an availability zone.
e.g.
``us-east-1a``
``us-east-1b``
Security Group
Instances belong to one or more security groups. A group is created for each security group, with all characters except alphanumerics, dashes (-) converted to underscores (_). Each group is prefixed by ``security_group_``
e.g.
``security_group_default``
``security_group_webservers``
``security_group_Pete_s_Fancy_Group``
Tags
Each instance can have a variety of key/value pairs associated with it called Tags. The most common tag key is 'Name', though anything is possible. Each key/value pair is its own group of instances, again with special characters converted to underscores, in the format ``tag_KEY_VALUE``
When the Ansible is interacting with a specific server, the EC2 inventory script is called again with the ``--host HOST`` option. This looks up the HOST in the index cache to get the instance ID, and then makes an API call to AWS to get information about that specific instance. It then makes information about that instance available as variables to your playbooks. Each variable is prefixed by ``ec2_``. Here are some of the variables available:
- ec2_architecture
- ec2_description
- ec2_dns_name
- ec2_id
- ec2_image_id
- ec2_instance_type
- ec2_ip_address
- ec2_kernel
- ec2_key_name
- ec2_launch_time
- ec2_monitored
- ec2_ownerId
- ec2_placement
- ec2_platform
- ec2_previous_state
- ec2_private_dns_name
- ec2_private_ip_address
- ec2_public_dns_name
- ec2_ramdisk
- ec2_region
- ec2_root_device_name
- ec2_root_device_type
- ec2_security_group_ids
- ec2_security_group_names
- ec2_spot_instance_request_id
- ec2_state
- ec2_state_code
- ec2_state_reason
- ec2_status
- ec2_subnet_id
- ec2_tag_Name
- ec2_tenancy
- ec2_virtualization_type
- ec2_vpc_id
Both ``ec2_security_group_ids`` and ``ec2_security_group_names`` are comma-separated lists of all security groups. Each EC2 tag is a variable in the format ``ec2_tag_KEY``.
To see the complete list of variables available for an instance, run the script by itself::
Though not detailed here in as much depth as the EC2 module, there's also a OpenStack Compute external inventory source in the plugins directory. It requires the Grizzly release of OpenStack or
later. See the inline comments in the module source for how to use it.
Callback Plugins
----------------
Ansible can be configured via code to respond to external events. This can include enhancing logging, signalling an external software
system, or even (yes, really) making sound effects. Some examples are contained in the plugins directory.
Connection Type Plugins
-----------------------
By default, ansible ships with a 'paramiko' SSH, native ssh (just called 'ssh'), and 'local' connection type, and an accelerated connection type named 'fireball'. All of these can be used
in playbooks and with /usr/bin/ansible to decide how you want to talk to remote machines. The basics of these connection types
are covered in the 'getting started' section. Should you want to extend Ansible to support other transports (SNMP? Message bus?
Carrier Pigeon?) it's as simple as copying the format of one of the existing modules and dropping it into the connection plugins
directory. The value of 'smart' for a connection allows selection of paramiko or openssh based on system capabilities, and chooses
'ssh' if OpenSSH supports ControlPersist, in Ansible 1.2.1 an later. Previous versions did not support 'smart'.
Lookup Plugins
--------------
Language constructs like "with_fileglob" and "with_items" are implemented via lookup plugins. Just like other plugin types, you can write your own.
Vars Plugins
------------
Playbook constructs like 'host_vars' and 'group_vars' work via 'vars' plugins. They inject additional variable
data into ansible runs that did not come from an inventory, playbook, or command line. Note that variables
can also be returned from inventory, so in most cases, you won't need to write or understand vars_plugins.
Filter Plugins
--------------
If you want more Jinja2 filters available in a Jinja2 template (filters like to_yaml and to_json are provided by default), they can be extended by writing a filter plugin.
Distributing Plugins
--------------------
.. versionadded:: 0.8
Plugins are loaded from both Python's site_packages (those that ship with ansible) and a configured plugins directory, which defaults
to /usr/share/ansible/plugins, in a subfolder for each plugin type::
* action_plugins
* lookup_plugins
* callback_plugins
* connection_plugins
* filter_plugins
* vars_plugins
To change this path, edit the ansible configuration file.
In addition, plugins can be shipped in a subdirectory relative to a top-level playbook, in folders named the same as indicated above.
.. seealso::
:doc:`modules`
List of built-in modules
`Mailing List <http://groups.google.com/group/ansible-project>`_
Questions? Help? Ideas? Stop by the list on Google Groups
for (hostname, result) in results['contacted'].items():
if not 'failed' in result:
print "%s >>> %s" % (hostname, result['stdout'])
print "FAILED *******"
for (hostname, result) in results['contacted'].items():
if 'failed' in result:
print "%s >>> %s" % (hostname, result['msg'])
print "DOWN *********"
for (hostname, result) in results['dark'].items():
print "%s >>> %s" % (hostname, result)
Advanced programmers may also wish to read the source to ansible itself, for
it uses the Runner() API (with all available options) to implement the
command line tools ``ansible`` and ``ansible-playbook``.
Plugins Available Online
------------------------
The remainder of features in the API docs have components available in `ansible-plugins <https://github.com/ansible/ansible/blob/devel/plugins>`_. Send us a github pull request if you develop any interesting features.
External Inventory Scripts
--------------------------
Often a user of a configuration management system will want to keep inventory
in a different system. Frequent examples include LDAP, `Cobbler <http://cobbler.github.com>`_,
or a piece of expensive enterprisey CMDB software. Ansible easily supports all
of these options via an external inventory system. The plugins directory contains some of these already -- including options for EC2/Eucalyptus and OpenStack, which will be detailed below.
It's possible to write an external inventory script in any language. If you are familiar with Puppet terminology, this concept is basically the same as 'external nodes', with the slight difference that it also defines which hosts are managed.
Simple! We just create a script that can return JSON in the right format when fed the proper arguments.
Script Conventions
``````````````````
When the external node script is called with the single argument '--list', the script must return a JSON hash/dictionary of all the groups to be managed.
Each group's value should be either a hash/dictionary containing a list of each host/IP, potential child groups, and potential group variables, or
simply a list of host/IP addresses, like so::
When the external node script is called with the single argument '--list', the script must return a JSON hash/dictionary of all the groups to be managed. Each group's value should be either a hash/dictionary containing a list of each host/IP, potential child groups, and potential group variables, or simply a list of host/IP addresses, like so::
{
"databases" : {
...
...
@@ -167,231 +76,3 @@ The data to be added to the top level JSON dictionary looks like this::
}
Example: The Cobbler External Inventory Script
``````````````````````````````````````````````
It is expected that many Ansible users will also be `Cobbler <http://cobbler.github.com>`_ users. Cobbler has a generic
layer that allows it to represent data for multiple configuration management systems (even at the same time), and has
been referred to as a 'lightweight CMDB' by some admins. This particular script will communicate with Cobbler
using Cobbler's XMLRPC API.
To tie Ansible's inventory to Cobbler (optional), copy `this script <https://raw.github.com/ansible/ansible/devel/plugins/inventory/cobbler.py>`_ to /etc/ansible/hosts and `chmod +x` the file. cobblerd will now need
to be running when you are using Ansible.
Test the file by running `./etc/ansible/hosts` directly. You should see some JSON data output, but it may not have
anything in it just yet.
Let's explore what this does. In cobbler, assume a scenario somewhat like the following::
cobbler system edit --name=foo --dns-name="foo.example.com" --mgmt-classes="atlanta" --ksmeta="c=4"
cobbler system edit --name=bar --dns-name="bar.example.com" --mgmt-classes="atlanta" --ksmeta="c=5"
In the example above, the system 'foo.example.com' will be addressable by ansible directly, but will also be addressable when using the group names 'webserver' or 'atlanta'. Since Ansible uses SSH, we'll try to contract system foo over 'foo.example.com', only, never just 'foo'. Similarly, if you try "ansible foo" it wouldn't find the system... but "ansible 'foo*'" would, because the system DNS name starts with 'foo'.
The script doesn't just provide host and group info. In addition, as a bonus, when the 'setup' module is run (which happens automatically when using playbooks), the variables 'a', 'b', and 'c' will all be auto-populated in the templates::
# file: /srv/motd.j2
Welcome, I am templated with a value of a={{ a }}, b={{ b }}, and c={{ c }}
Which could be executed just like this::
ansible webserver -m setup
ansible webserver -m template -a "src=/tmp/motd.j2 dest=/etc/motd"
.. note::
The name 'webserver' came from cobbler, as did the variables for
the config file. You can still pass in your own variables like
normal in Ansible, but variables from the external inventory script
will override any that have the same name.
So, with the template above (motd.j2), this would result in the following data being written to /etc/motd for system 'foo'::
Welcome, I am templated with a value of a=2, b=3, and c=4
And on system 'bar' (bar.example.com)::
Welcome, I am templated with a value of a=2, b=3, and c=5
And technically, though there is no major good reason to do it, this also works too::
ansible webserver -m shell -a "echo {{ a }}"
So in other words, you can use those variables in arguments/actions as well. You might use this to name
a conf.d file appropriately or something similar. Who knows?
So that's the Cobbler integration support -- using the cobbler script as an example, it should be trivial to adapt Ansible to pull inventory, as well as variable information, from any data source. If you create anything interesting, please share with the mailing list, and we can keep it in the source code tree for others to use.
Example: AWS EC2 External Inventory Script
``````````````````````````````````````````
If you use Amazon Web Services EC2, maintaining an inventory file might not be the best approach. For this reason, you can use the `EC2 external inventory <https://raw.github.com/ansible/ansible/devel/plugins/inventory/ec2.py>`_ script.
You can use this script in one of two ways. The easiest is to use Ansible's ``-i`` command line option and specify the path to the script.
ansible -i ec2.py -u ubuntu us-east-1d -m ping
The second option is to copy the script to `/etc/ansible/hosts` and `chmod +x` it. You will also need to copy the `ec2.ini <https://raw.github.com/ansible/ansible/devel/plugins/inventory/ec2.ini>`_ file to `/etc/ansible/ec2.ini`. Then you can run ansible as you would normally.
To successfully make an API call to AWS, you will need to configure Boto (the Python interface to AWS). There are a `variety of methods <http://docs.pythonboto.org/en/latest/boto_config_tut.html>`_ available, but the simplest is just to export two environment variables:
export AWS_ACCESS_KEY_ID='AK123'
export AWS_SECRET_ACCESS_KEY='abc123'
You can test the script by itself to make sure your config is correct
cd plugins/inventory
./ec2.py --list
After a few moments, you should see your entire EC2 inventory across all regions in JSON.
Since each region requires its own API call, if you are only using a small set of regions, feel free to edit ``ec2.ini`` and list only the regions you are interested in. There are other config options in ``ec2.ini`` including cache control, and destination variables.
At their heart, inventory files are simply a mapping from some name to a destination address. The default ``ec2.ini`` settings are configured for running Ansible from outside EC2 (from your laptop for example). If you are running Ansible from within EC2, internal DNS names and IP addresses may make more sense than public DNS names. In this case, you can modify the ``destination_variable`` in ``ec2.ini`` to be the private DNS name of an instance. This is particularly important when running Ansible within a private subnet inside a VPC, where the only way to access an instance is via its private IP address. For VPC instances, `vpc_destination_variable` in ``ec2.ini`` provides a means of using which ever `boto.ec2.instance variable <http://docs.pythonboto.org/en/latest/ref/ec2.html#module-boto.ec2.instance>`_ makes the most sense for your use case.
The EC2 external inventory provides mappings to instances from several groups:
Instance ID
These are groups of one since instance IDs are unique.
e.g.
``i-00112233``
``i-a1b1c1d1``
Region
A group of all instances in an AWS region.
e.g.
``us-east-1``
``us-west-2``
Availability Zone
A group of all instances in an availability zone.
e.g.
``us-east-1a``
``us-east-1b``
Security Group
Instances belong to one or more security groups. A group is created for each security group, with all characters except alphanumerics, dashes (-) converted to underscores (_). Each group is prefixed by ``security_group_``
e.g.
``security_group_default``
``security_group_webservers``
``security_group_Pete_s_Fancy_Group``
Tags
Each instance can have a variety of key/value pairs associated with it called Tags. The most common tag key is 'Name', though anything is possible. Each key/value pair is its own group of instances, again with special characters converted to underscores, in the format ``tag_KEY_VALUE``
When the Ansible is interacting with a specific server, the EC2 inventory script is called again with the ``--host HOST`` option. This looks up the HOST in the index cache to get the instance ID, and then makes an API call to AWS to get information about that specific instance. It then makes information about that instance available as variables to your playbooks. Each variable is prefixed by ``ec2_``. Here are some of the variables available:
- ec2_architecture
- ec2_description
- ec2_dns_name
- ec2_id
- ec2_image_id
- ec2_instance_type
- ec2_ip_address
- ec2_kernel
- ec2_key_name
- ec2_launch_time
- ec2_monitored
- ec2_ownerId
- ec2_placement
- ec2_platform
- ec2_previous_state
- ec2_private_dns_name
- ec2_private_ip_address
- ec2_public_dns_name
- ec2_ramdisk
- ec2_region
- ec2_root_device_name
- ec2_root_device_type
- ec2_security_group_ids
- ec2_security_group_names
- ec2_spot_instance_request_id
- ec2_state
- ec2_state_code
- ec2_state_reason
- ec2_status
- ec2_subnet_id
- ec2_tag_Name
- ec2_tenancy
- ec2_virtualization_type
- ec2_vpc_id
Both ``ec2_security_group_ids`` and ``ec2_security_group_names`` are comma-separated lists of all security groups. Each EC2 tag is a variable in the format ``ec2_tag_KEY``.
To see the complete list of variables available for an instance, run the script by itself::
Though not detailed here in as much depth as the EC2 module, there's also a OpenStack Compute external inventory source in the plugins directory. It requires the Grizzly release of OpenStack or
later. See the inline comments in the module source for how to use it.
Callback Plugins
----------------
Ansible can be configured via code to respond to external events. This can include enhancing logging, signalling an external software
system, or even (yes, really) making sound effects. Some examples are contained in the plugins directory.
Connection Type Plugins
-----------------------
By default, ansible ships with a 'paramiko' SSH, native ssh (just called 'ssh'), and 'local' connection type, and an accelerated connection type named 'fireball'. All of these can be used
in playbooks and with /usr/bin/ansible to decide how you want to talk to remote machines. The basics of these connection types
are covered in the 'getting started' section. Should you want to extend Ansible to support other transports (SNMP? Message bus?
Carrier Pigeon?) it's as simple as copying the format of one of the existing modules and dropping it into the connection plugins
directory. The value of 'smart' for a connection allows selection of paramiko or openssh based on system capabilities, and chooses
'ssh' if OpenSSH supports ControlPersist, in Ansible 1.2.1 an later. Previous versions did not support 'smart'.
Lookup Plugins
--------------
Language constructs like "with_fileglob" and "with_items" are implemented via lookup plugins. Just like other plugin types, you can write your own.
Vars Plugins
------------
Playbook constructs like 'host_vars' and 'group_vars' work via 'vars' plugins. They inject additional variable
data into ansible runs that did not come from an inventory, playbook, or command line. Note that variables
can also be returned from inventory, so in most cases, you won't need to write or understand vars_plugins.
Filter Plugins
--------------
If you want more Jinja2 filters available in a Jinja2 template (filters like to_yaml and to_json are provided by default), they can be extended by writing a filter plugin.
Distributing Plugins
--------------------
.. versionadded:: 0.8
Plugins are loaded from both Python's site_packages (those that ship with ansible) and a configured plugins directory, which defaults
to /usr/share/ansible/plugins, in a subfolder for each plugin type::
* action_plugins
* lookup_plugins
* callback_plugins
* connection_plugins
* filter_plugins
* vars_plugins
To change this path, edit the ansible configuration file.
In addition, plugins can be shipped in a subdirectory relative to a top-level playbook, in folders named the same as indicated above.
.. seealso::
:doc:`modules`
List of built-in modules
`Mailing List <http://groups.google.com/group/ansible-project>`_
Questions? Help? Ideas? Stop by the list on Google Groups
for (hostname, result) in results['contacted'].items():
if not 'failed' in result:
print "%s >>> %s" % (hostname, result['stdout'])
print "FAILED *******"
for (hostname, result) in results['contacted'].items():
if 'failed' in result:
print "%s >>> %s" % (hostname, result['msg'])
print "DOWN *********"
for (hostname, result) in results['dark'].items():
print "%s >>> %s" % (hostname, result)
Advanced programmers may also wish to read the source to ansible itself, for
it uses the Runner() API (with all available options) to implement the
command line tools ``ansible`` and ``ansible-playbook``.
Plugins Available Online
------------------------
The remainder of features in the API docs have components available in `ansible-plugins <https://github.com/ansible/ansible/blob/devel/plugins>`_. Send us a github pull request if you develop any interesting features.
External Inventory Scripts
--------------------------
Often a user of a configuration management system will want to keep inventory
in a different system. Frequent examples include LDAP, `Cobbler <http://cobbler.github.com>`_,
or a piece of expensive enterprisey CMDB software. Ansible easily supports all
of these options via an external inventory system. The plugins directory contains some of these already -- including options for EC2/Eucalyptus and OpenStack, which will be detailed below.
It's possible to write an external inventory script in any language. If you are familiar with Puppet terminology, this concept is basically the same as 'external nodes', with the slight difference that it also defines which hosts are managed.
Script Conventions
``````````````````
When the external node script is called with the single argument '--list', the script must return a JSON hash/dictionary of all the groups to be managed.
Each group's value should be either a hash/dictionary containing a list of each host/IP, potential child groups, and potential group variables, or
Before version 1.0, each group could only have a list of hostnames/IP addresses, like the webservers, marietta, and 5points groups above.
When called with the arguments '--host <hostname>' (where <hostname> is a host from above), the script must return either an empty JSON
hash/dictionary, or a hash/dictionary of variables to make available to templates and playbooks. Returning variables is optional,
if the script does not wish to do this, returning an empty hash/dictionary is the way to go::
{
"favcolor" : "red",
"ntpserver" : "wolf.example.com",
"monitoring" : "pack.example.com"
}
Tuning the External Inventory Script
````````````````````````````````````
.. versionadded:: 1.3
The stock inventory script system detailed above works for all versions of Ansible, but calling
'--host' for every host can be rather expensive, especially if it involves expensive API calls to
a remote subsystemm. In Ansible
1.3 or later, if the inventory script returns a top level element called "_meta", it is possible
to return all of the host variables in one inventory script call. When this meta element contains
a value for "hostvars", the inventory script will not be invoked with "--host" for each host. This
results in a significant performance increase for large numbers of hosts, and also makes client
side caching easier to implement for the inventory script.
The data to be added to the top level JSON dictionary looks like this::
{
# results of inventory script as above go here
# ...
"_meta" : {
"hostvars" : {
"moocow.example.com" : { "asdf" : 1234 },
"llama.example.com" : { "asdf" : 5678 },
}
}
}
Example: The Cobbler External Inventory Script
``````````````````````````````````````````````
It is expected that many Ansible users will also be `Cobbler <http://cobbler.github.com>`_ users. Cobbler has a generic
layer that allows it to represent data for multiple configuration management systems (even at the same time), and has
been referred to as a 'lightweight CMDB' by some admins. This particular script will communicate with Cobbler
using Cobbler's XMLRPC API.
To tie Ansible's inventory to Cobbler (optional), copy `this script <https://raw.github.com/ansible/ansible/devel/plugins/inventory/cobbler.py>`_ to /etc/ansible/hosts and `chmod +x` the file. cobblerd will now need
to be running when you are using Ansible.
Test the file by running `./etc/ansible/hosts` directly. You should see some JSON data output, but it may not have
anything in it just yet.
Let's explore what this does. In cobbler, assume a scenario somewhat like the following::
cobbler system edit --name=foo --dns-name="foo.example.com" --mgmt-classes="atlanta" --ksmeta="c=4"
cobbler system edit --name=bar --dns-name="bar.example.com" --mgmt-classes="atlanta" --ksmeta="c=5"
In the example above, the system 'foo.example.com' will be addressable by ansible directly, but will also be addressable when using the group names 'webserver' or 'atlanta'. Since Ansible uses SSH, we'll try to contract system foo over 'foo.example.com', only, never just 'foo'. Similarly, if you try "ansible foo" it wouldn't find the system... but "ansible 'foo*'" would, because the system DNS name starts with 'foo'.
The script doesn't just provide host and group info. In addition, as a bonus, when the 'setup' module is run (which happens automatically when using playbooks), the variables 'a', 'b', and 'c' will all be auto-populated in the templates::
# file: /srv/motd.j2
Welcome, I am templated with a value of a={{ a }}, b={{ b }}, and c={{ c }}
Which could be executed just like this::
ansible webserver -m setup
ansible webserver -m template -a "src=/tmp/motd.j2 dest=/etc/motd"
.. note::
The name 'webserver' came from cobbler, as did the variables for
the config file. You can still pass in your own variables like
normal in Ansible, but variables from the external inventory script
will override any that have the same name.
So, with the template above (motd.j2), this would result in the following data being written to /etc/motd for system 'foo'::
Welcome, I am templated with a value of a=2, b=3, and c=4
And on system 'bar' (bar.example.com)::
Welcome, I am templated with a value of a=2, b=3, and c=5
And technically, though there is no major good reason to do it, this also works too::
ansible webserver -m shell -a "echo {{ a }}"
So in other words, you can use those variables in arguments/actions as well. You might use this to name
a conf.d file appropriately or something similar. Who knows?
So that's the Cobbler integration support -- using the cobbler script as an example, it should be trivial to adapt Ansible to pull inventory, as well as variable information, from any data source. If you create anything interesting, please share with the mailing list, and we can keep it in the source code tree for others to use.
Example: AWS EC2 External Inventory Script
``````````````````````````````````````````
If you use Amazon Web Services EC2, maintaining an inventory file might not be the best approach. For this reason, you can use the `EC2 external inventory <https://raw.github.com/ansible/ansible/devel/plugins/inventory/ec2.py>`_ script.
You can use this script in one of two ways. The easiest is to use Ansible's ``-i`` command line option and specify the path to the script.
ansible -i ec2.py -u ubuntu us-east-1d -m ping
The second option is to copy the script to `/etc/ansible/hosts` and `chmod +x` it. You will also need to copy the `ec2.ini <https://raw.github.com/ansible/ansible/devel/plugins/inventory/ec2.ini>`_ file to `/etc/ansible/ec2.ini`. Then you can run ansible as you would normally.
To successfully make an API call to AWS, you will need to configure Boto (the Python interface to AWS). There are a `variety of methods <http://docs.pythonboto.org/en/latest/boto_config_tut.html>`_ available, but the simplest is just to export two environment variables:
export AWS_ACCESS_KEY_ID='AK123'
export AWS_SECRET_ACCESS_KEY='abc123'
You can test the script by itself to make sure your config is correct
cd plugins/inventory
./ec2.py --list
After a few moments, you should see your entire EC2 inventory across all regions in JSON.
Since each region requires its own API call, if you are only using a small set of regions, feel free to edit ``ec2.ini`` and list only the regions you are interested in. There are other config options in ``ec2.ini`` including cache control, and destination variables.
At their heart, inventory files are simply a mapping from some name to a destination address. The default ``ec2.ini`` settings are configured for running Ansible from outside EC2 (from your laptop for example). If you are running Ansible from within EC2, internal DNS names and IP addresses may make more sense than public DNS names. In this case, you can modify the ``destination_variable`` in ``ec2.ini`` to be the private DNS name of an instance. This is particularly important when running Ansible within a private subnet inside a VPC, where the only way to access an instance is via its private IP address. For VPC instances, `vpc_destination_variable` in ``ec2.ini`` provides a means of using which ever `boto.ec2.instance variable <http://docs.pythonboto.org/en/latest/ref/ec2.html#module-boto.ec2.instance>`_ makes the most sense for your use case.
The EC2 external inventory provides mappings to instances from several groups:
Instance ID
These are groups of one since instance IDs are unique.
e.g.
``i-00112233``
``i-a1b1c1d1``
Region
A group of all instances in an AWS region.
e.g.
``us-east-1``
``us-west-2``
Availability Zone
A group of all instances in an availability zone.
e.g.
``us-east-1a``
``us-east-1b``
Security Group
Instances belong to one or more security groups. A group is created for each security group, with all characters except alphanumerics, dashes (-) converted to underscores (_). Each group is prefixed by ``security_group_``
e.g.
``security_group_default``
``security_group_webservers``
``security_group_Pete_s_Fancy_Group``
Tags
Each instance can have a variety of key/value pairs associated with it called Tags. The most common tag key is 'Name', though anything is possible. Each key/value pair is its own group of instances, again with special characters converted to underscores, in the format ``tag_KEY_VALUE``
When the Ansible is interacting with a specific server, the EC2 inventory script is called again with the ``--host HOST`` option. This looks up the HOST in the index cache to get the instance ID, and then makes an API call to AWS to get information about that specific instance. It then makes information about that instance available as variables to your playbooks. Each variable is prefixed by ``ec2_``. Here are some of the variables available:
- ec2_architecture
- ec2_description
- ec2_dns_name
- ec2_id
- ec2_image_id
- ec2_instance_type
- ec2_ip_address
- ec2_kernel
- ec2_key_name
- ec2_launch_time
- ec2_monitored
- ec2_ownerId
- ec2_placement
- ec2_platform
- ec2_previous_state
- ec2_private_dns_name
- ec2_private_ip_address
- ec2_public_dns_name
- ec2_ramdisk
- ec2_region
- ec2_root_device_name
- ec2_root_device_type
- ec2_security_group_ids
- ec2_security_group_names
- ec2_spot_instance_request_id
- ec2_state
- ec2_state_code
- ec2_state_reason
- ec2_status
- ec2_subnet_id
- ec2_tag_Name
- ec2_tenancy
- ec2_virtualization_type
- ec2_vpc_id
Both ``ec2_security_group_ids`` and ``ec2_security_group_names`` are comma-separated lists of all security groups. Each EC2 tag is a variable in the format ``ec2_tag_KEY``.
To see the complete list of variables available for an instance, run the script by itself::
Though not detailed here in as much depth as the EC2 module, there's also a OpenStack Compute external inventory source in the plugins directory. It requires the Grizzly release of OpenStack or
later. See the inline comments in the module source for how to use it.
Callback Plugins
----------------
Ansible can be configured via code to respond to external events. This can include enhancing logging, signalling an external software
system, or even (yes, really) making sound effects. Some examples are contained in the plugins directory.
This section will explore these features, though they are generally not common in terms of things people would look to extend.
Connection Type Plugins
-----------------------
By default, ansible ships with a 'paramiko' SSH, native ssh (just called 'ssh'), and 'local' connection type, and an accelerated connection type named 'fireball'. All of these can be used
By default, ansible ships with a 'paramiko' SSH, native ssh (just called 'ssh'), and 'local' connection type, and an accelerated connection type named 'fireball' -- there are also some minor players like 'chroot' and 'jail'. All of these can be used
in playbooks and with /usr/bin/ansible to decide how you want to talk to remote machines. The basics of these connection types
are covered in the 'getting started' section. Should you want to extend Ansible to support other transports (SNMP? Message bus?
Carrier Pigeon?) it's as simple as copying the format of one of the existing modules and dropping it into the connection plugins
directory. The value of 'smart' for a connection allows selection of paramiko or openssh based on system capabilities, and chooses
'ssh' if OpenSSH supports ControlPersist, in Ansible 1.2.1 an later. Previous versions did not support 'smart'.
More documentation on writing connection plugins is pending, though you can jump into lib/ansible/runner/connection_plugins and figure
things out pretty easily.
Lookup Plugins
--------------
Language constructs like "with_fileglob" and "with_items" are implemented via lookup plugins. Just like other plugin types, you can write your own.
More documentation on writing connection plugins is pending, though you can jump into lib/ansible/runner/lookup_plugins and figure
things out pretty easily.
Vars Plugins
------------
...
...
@@ -362,10 +35,46 @@ Playbook constructs like 'host_vars' and 'group_vars' work via 'vars' plugins.
data into ansible runs that did not come from an inventory, playbook, or command line. Note that variables
can also be returned from inventory, so in most cases, you won't need to write or understand vars_plugins.
More documentation on writing connection plugins is pending, though you can jump into lib/ansible/inventory/vars_plugins and figure
things out pretty easily.
If you find yourself wanting to write a vars_plugin, it's more likely you should write an inventory script instead.
Filter Plugins
--------------
If you want more Jinja2 filters available in a Jinja2 template (filters like to_yaml and to_json are provided by default), they can be extended by writing a filter plugin.
If you want more Jinja2 filters available in a Jinja2 template (filters like to_yaml and to_json are provided by default), they can be extended by writing a filter plugin. Most of the time, when someone comes up with an idea for a new filter they would like to make available in a playbook, we'll just include them in 'core.py' instead.
Jump into lib/ansible/runner/filter_plugins/ for details.
Callbacks
---------
Callbacks are one of the more interesting plugin types. Adding additional callback plugins to Ansible allows for adding new behaviors when responding to events.
Examples
++++++++
Example callbacks are shown `in github in the callbacks directory <https://github.com/ansible/ansible/tree/devel/plugins/callbacks>_`.
The 'log_plays' callback is an example of how to intercept playbook events to a log file, and the 'mail' callback sends email
when playbooks complete.
The 'osx_say' callback provided is particularly entertaining -- it will respond with computer synthesized speech on OS X in relation
to playbook events, and is guaranteed to entertain and/or annoy coworkers.
Configuring
+++++++++++
To active a callback drop it in a callback directory as configured in ansible.cfg.
Development
+++++++++++
More information will come later, though see the source of any of the existing callbacks and you should be able to get started quickly.
The variable 'ansible_os_family' is being interpolated into
the list of filenames being defined for vars_files.
As a reminder, the various YAML files contain just keys and values::
---
# for vars/CentOS.yml
apache: httpd
somethingelse: 42
How does this work? If the operating system was 'CentOS', the first file Ansible would try to import
would be 'vars/CentOS.yml', followed by '/vars/os_defaults.yml' if that file
did not exist. If no files in the list were found, an error would be raised.
On Debian, it would instead first look towards 'vars/Debian.yml' instead of 'vars/CentOS.yml', before
falling back on 'vars/os_defaults.yml'. Pretty simple.
To use this conditional import feature, you'll need facter or ohai installed prior to running the playbook, but
you can of course push this out with Ansible if you like::
# for facter
ansible -m yum -a "pkg=facter ensure=installed"
ansible -m yum -a "pkg=ruby-json ensure=installed"
# for ohai
ansible -m yum -a "pkg=ohai ensure=installed"
Ansible's approach to configuration -- separating variables from tasks, keeps your playbooks
from turning into arbitrary code with ugly nested ifs, conditionals, and so on - and results
in more streamlined & auditable configuration rules -- especially because there are a
minimum of decision points to track.
Selecting Files And Templates Based On Variables
````````````````````````````````````````````````
.. note:: This is an advanced topic that is infrequently used. You can probably skip this section.
Sometimes a configuration file you want to copy, or a template you will use may depend on a variable.
The following construct selects the first available file appropriate for the variables of a given host, which is often much cleaner than putting a lot of if conditionals in a template.
The following example shows how to template out a configuration file that was very different between, say, CentOS and Debian::
- name: template a file
template: src={{ item }} dest=/etc/myapp/foo.conf
with_first_found:
files:
- {{ ansible_distribution }}.conf
- default.conf
paths:
- search_location_one/somedir/
- /opt/other_location/somedir/
Register Variables
``````````````````
.. versionadded:: 0.7
Often in a playbook it may be useful to store the result of a given command in a variable and access
it later. Use of the command module in this way can in many ways eliminate the need to write site specific facts, for
instance, you could test for the existence of a particular program.
The 'register' keyword decides what variable to save a result in. The resulting variables can be used in templates, action lines, or *when* statements. It looks like this (in an obviously trivial example)::
- name: test play
hosts: all
tasks:
- shell: cat /etc/motd
register: motd_contents
- shell: echo "motd contains the word hi"
when: motd_contents.stdout.find('hi') != -1
As shown previously, the registered variable's string contents are accessible with the 'stdout' value.
The registered result can be used in the "with_items" of a task if it is converted into
a list (or already is a list) as shown below. "stdout_lines" is already available on the object as
well though you could also call "home_dirs.stdout.split()" if you wanted, and could split by other
fields::
- name: registered variable usage as a with_items list
Additionally, *group_names* is a list (array) of all the groups the current host is in. This can be used in templates using Jinja2 syntax to make template source files that vary based on the group membership (or role) of the host::
{% if 'webserver' in group_names %}
# some part of a configuration file that only applies to webservers
{% endif %}
*groups* is a list of all the groups (and hosts) in the inventory. This can be used to enumerate all hosts within a group.
For example::
{% for host in groups['app_servers'] %}
# something that applies to all app servers.
{% endfor %}
A frequently used idiom is walking a group to find all IP addresses in that group::
An example of this could include pointing a frontend proxy server to all of the app servers, setting up the correct firewall rules between servers, etc.
Just a few other 'magic' variables are available... There aren't many.
Additionally, *inventory_hostname* is the name of the hostname as configured in Ansible's inventory host file. This can
be useful for when you don't want to rely on the discovered hostname `ansible_hostname` or for other mysterious
reasons. If you have a long FQDN, *inventory_hostname_short* also contains the part up to the first
period, without the rest of the domain.
Don't worry about any of this unless you think you need it. You'll know when you do.
Also available, *inventory_dir* is the pathname of the directory holding Ansible's inventory host file, *inventory_file* is the pathname and the filename pointing to the Ansible's inventory host file.
Variable File Separation
````````````````````````
It's a great idea to keep your playbooks under source control, but
you may wish to make the playbook source public while keeping certain
important variables private. Similarly, sometimes you may just
want to keep certain information in different files, away from
the main playbook.
You can do this by using an external variables file, or files, just like this::
---
- hosts: all
remote_user: root
vars:
favcolor: blue
vars_files:
- /vars/external_vars.yml
tasks:
- name: this is just a placeholder
command: /bin/echo foo
This removes the risk of sharing sensitive data with others when
sharing your playbook source with them.
The contents of each variables file is a simple YAML dictionary, like this::
---
# in the above example, this would be vars/external_vars.yml
somevar: somevalue
password: magic
.. note::
It's also possible to keep per-host and per-group variables in very
similar files, this is covered in :ref:`patterns`.
Prompting For Sensitive Data
````````````````````````````
You may wish to prompt the user for certain input, and can
do so with the similarly named 'vars_prompt' section. This has uses
beyond security, for instance, you may use the same playbook for all
software releases and would prompt for a particular release version
in a push-script::
---
- hosts: all
remote_user: root
vars:
from: "camelot"
vars_prompt:
name: "what is your name?"
quest: "what is your quest?"
favcolor: "what is your favorite color?"
There are full examples of both of these items in the github examples/playbooks directory.
If you have a variable that changes infrequently, it might make sense to
provide a default value that can be overridden. This can be accomplished using
the default argument::
vars_prompt:
- name: "release_version"
prompt: "Product release version"
default: "1.0"
An alternative form of vars_prompt allows for hiding input from the user, and may later support
some other options, but otherwise works equivalently::
vars_prompt:
- name: "some_password"
prompt: "Enter password"
private: yes
- name: "release_version"
prompt: "Product release version"
private: no
If `Passlib <http://pythonhosted.org/passlib/>`_ is installed, vars_prompt can also crypt the
entered value so you can use it, for instance, with the user module to define a password::
vars_prompt:
- name: "my_password2"
prompt: "Enter password2"
private: yes
encrypt: "md5_crypt"
confirm: yes
salt_size: 7
You can use any crypt scheme supported by 'Passlib':
As with the case of 'with_items' above, you can use previously defined variables. Just specify the variable's name without templating it with '{{ }}'::
- name: here, 'users' contains the above list of employees
``with_password`` and associated lookup macro generate a random plaintext password and store it in
a file at a given filepath. Support for crypted save modes (as with vars_prompt) is pending. If the
file exists previously, it will retrieve its contents, behaving just like with_file. Usage of variables like "{{ inventory_hostname }}" in the filepath can be used to set
up random passwords per host (what simplifies password management in 'host_vars' variables).
Generated passwords contain a random mix of upper and lowercase ASCII letters, the
numbers 0-9 and punctuation (". , : - _"). The default length of a generated password is 30 characters.
This length can be changed by passing an extra parameter::
It is quite possible that you may need to get package updates through a proxy, or even get some package
updates through a proxy and access other packages not through a proxy. Ansible makes it easy for you
to configure your environment by using the 'environment' keyword. Here is an example::
- hosts: all
remote_user: root
tasks:
- apt: name=cobbler state=installed
environment:
http_proxy: http://proxy.example.com:8080
The environment can also be stored in a variable, and accessed like so::
- hosts: all
remote_user: root
# here we make a variable named "env" that is a dictionary
vars:
proxy_env:
http_proxy: http://proxy.example.com:8080
tasks:
- apt: name=cobbler state=installed
environment: "{{ proxy_env }}"
While just proxy settings were shown above, any number of settings can be supplied. The most logical place
to define an environment hash might be a group_vars file, like so::
---
# file: group_vars/boston
ntp_server: ntp.bos.example.com
backup: bak.bos.example.com
proxy_env:
http_proxy: http://proxy.bos.example.com:8080
https_proxy: http://proxy.bos.example.com:8080
Getting values from files
`````````````````````````
.. versionadded:: 0.8
Sometimes you'll want to include the content of a file directly into a playbook. You can do so using a macro.
This syntax will remain in future versions, though we will also will provide ways to do this via lookup plugins (see "More Loops") as well. What follows
is an example using the authorized_key module, which requires the actual text of the SSH key as a parameter::
Sometimes a configuration file you want to copy, or a template you will use may depend on a variable.
The following construct selects the first available file appropriate for the variables of a given host, which is often much cleaner than putting a lot of if conditionals in a template.
The following example shows how to template out a configuration file that was very different between, say, CentOS and Debian::
first_available_file is only available to the copy and template modules.
Asynchronous Actions and Polling
````````````````````````````````
By default tasks in playbooks block, meaning the connections stay open
until the task is done on each node. If executing playbooks with
a small parallelism value (aka ``--forks``), you may wish that long
running operations can go faster. The easiest way to do this is
to kick them off all at once and then poll until they are done.
You will also want to use asynchronous mode on very long running
operations that might be subject to timeout.
To launch a task asynchronously, specify its maximum runtime
and how frequently you would like to poll for status. The default
poll value is 10 seconds if you do not specify a value for `poll`::
---
- hosts: all
remote_user: root
tasks:
- name: simulate long running op (15 sec), wait for up to 45, poll every 5
command: /bin/sleep 15
async: 45
poll: 5
.. note::
There is no default for the async time limit. If you leave off the
'async' keyword, the task runs synchronously, which is Ansible's
default.
Alternatively, if you do not need to wait on the task to complete, you may
"fire and forget" by specifying a poll value of 0::
---
- hosts: all
remote_user: root
tasks:
- name: simulate long running op, allow to run for 45, fire and forget
command: /bin/sleep 15
async: 45
poll: 0
.. note::
You shouldn't "fire and forget" with operations that require
exclusive locks, such as yum transactions, if you expect to run other
commands later in the playbook against those same resources.
.. note::
Using a higher value for ``--forks`` will result in kicking off asynchronous
tasks even faster. This also increases the efficiency of polling.
Local Playbooks
```````````````
It may be useful to use a playbook locally, rather than by connecting over SSH. This can be useful
for assuring the configuration of a system by putting a playbook on a crontab. This may also be used
to run a playbook inside a OS installer, such as an Anaconda kickstart.
To run an entire playbook locally, just set the "hosts:" line to "hosts:127.0.0.1" and then run the playbook like so::
ansible-playbook playbook.yml --connection=local
Alternatively, a local connection can be used in a single playbook play, even if other plays in the playbook
use the default remote connection type::
hosts: 127.0.0.1
connection: local
Turning Off Facts
`````````````````
If you know you don't need any fact data about your hosts, and know everything about your systems centrally, you
can turn off fact gathering. This has advantages in scaling Ansible in push mode with very large numbers of
systems, mainly, or if you are using Ansible on experimental platforms. In any play, just do this::
- hosts: whatever
gather_facts: no
Pull-Mode Playbooks
```````````````````
The use of playbooks in local mode (above) is made extremely powerful with the addition of `ansible-pull`.
A script for setting up ansible-pull is provided in the examples/playbooks directory of the source
checkout.
The basic idea is to use Ansible to set up a remote copy of Ansible on each managed node, each set to run via
cron and update playbook source via git. This inverts the default push architecture of Ansible into a pull
architecture, which has near-limitless scaling potential. The setup playbook can be tuned to change
the cron frequency, logging locations, and parameters to ansible-pull.
This is useful both for extreme scale-out as well as periodic remediation. Usage of the 'fetch' module to retrieve
logs from ansible-pull runs would be an excellent way to gather and analyze remote logs from ansible-pull.
Register Variables
``````````````````
.. versionadded:: 0.7
Often in a playbook it may be useful to store the result of a given command in a variable and access
it later. Use of the command module in this way can in many ways eliminate the need to write site specific facts, for
instance, you could test for the existence of a particular program.
The 'register' keyword decides what variable to save a result in. The resulting variables can be used in templates, action lines, or *when* statements. It looks like this (in an obviously trivial example)::
- name: test play
hosts: all
tasks:
- shell: cat /etc/motd
register: motd_contents
- shell: echo "motd contains the word hi"
when: motd_contents.stdout.find('hi') != -1
As shown previously, the registered variable's string contents are accessible with the 'stdout' value.
The registered result can be used in the "with_items" of a task if it is converted into
a list (or already is a list) as shown below. "stdout_lines" is already available on the object as
well though you could also call "home_dirs.stdout.split()" if you wanted, and could split by other
fields::
- name: registered variable usage as a with_items list
A common pattern is to use a local action to call 'rsync' to recursively copy files to the managed servers.
Here is an example::
---
# ...
tasks:
- name: recursively copy files from management server to target
local_action: command rsync -a /path/to/files {{ inventory_hostname }}:/path/to/target/
Note that you must have passphrase-less SSH keys or an ssh-agent configured for this to work, otherwise rsync
will need to ask for a passphrase.
Accelerated Mode
````````````````
.. versionadded:: 1.3
While SSH using the ControlPersist feature is quite fast and scalable, there is a certain amount of overhead involved in
creating connections. This can become something of a bottleneck when the number of hosts grows into the hundreds or
thousands. To help overcome this, Ansible offers an accelerated connection option. Accelerated mode can be anywhere from
2-6x faster than SSH with ControlPersist enabled, and 10x faster than paramiko.
Accelerated mode works by launching a temporary daemon over SSH. Once the daemon is running, Ansible will connect directly
to it via a raw socket connection. Ansible secures this communication by using a temporary AES key that is uploaded during
the SSH connection (this key is different for every host, and is also regenerated every time the daemon is started). By default,
Ansible will use port 5099 for the accelerated connection, though this is configurable. Once running, the daemon will accept
connections for 30 minutes, after which time it will terminate itself and need to be restarted over SSH.
Accelerated mode offers several improvments over the original fireball mode:
* No bootstrapping is required, only a single line needs to be added to each play you wish to run in accelerated mode.
* Support for sudo commands (see below for more details and caveats).
* Fewer requirements! ZeroMQ is no longer required, nor are there any special packages beyond python-keyczar.
In order to use accelerated mode, simply add `accelerate: true` to your play::
---
- hosts: all
accelerate: true
tasks:
- name: some task
command: echo {{ item }}
with_items:
- foo
- bar
- baz
If you wish to change the port Ansible will use for the accelerated connection, just add the `accelerated_port` option::
---
- hosts: all
accelerate: true
# default port is 5099
accelerate_port: 10000
The `accelerate_port` option can also be specified in the environment variable ACCELERATE_PORT, or in your `ansible.cfg` configuration::
[accelerate]
accelerate_port = 5099
As noted above, accelerated mode also supports running tasks via sudo, however there are two important caveats:
* You must remove requiretty from your sudoers options.
* Prompting for the sudo password is not yet supported, so the NOPASSWD option is required for commands.
Fireball Mode
`````````````
.. versionadded:: 0.8 (deprecated as of 1.3)
.. note::
The following section has been deprecated as of Ansible 1.3 in favor of the accelerated mode described above. This
documentation is here for users who may still be using the original fireball connection method only, and should not
be used for any new deployments.
Ansible's core connection types of 'local', 'paramiko', and 'ssh' are augmented in version 0.8 and later by a new extra-fast
connection type called 'fireball'. It can only be used with playbooks and does require some additional setup
outside the lines of Ansible's normal "no bootstrapping" philosophy. You are not required to use fireball mode
to use Ansible, though some users may appreciate it.
Fireball mode works by launching a temporary 0mq daemon from SSH that by default lives for only 30 minutes before
shutting off. Fireball mode, once running, uses temporary AES keys to encrypt a session, and requires direct
communication to given nodes on the configured port. The default is 5099. The fireball daemon runs as any user you
set it down as. So it can run as you, root, or so on. If multiple users are running Ansible as the same batch of hosts,
take care to use unique ports.
Fireball mode is roughly 10 times faster than paramiko for communicating with nodes and may be a good option
if you have a large number of hosts::
---
# set up the fireball transport
- hosts: all
gather_facts: no
connection: ssh # or paramiko
sudo: yes
tasks:
- action: fireball
# these operations will occur over the fireball transport
- hosts: all
connection: fireball
tasks:
- shell: echo "Hello {{ item }}"
with_items:
- one
- two
In order to use fireball mode, certain dependencies must be installed on both ends. You can use this playbook as a basis for initial bootstrapping on
any platform. You will also need gcc and zeromq-devel installed from your package manager, which you can of course also get Ansible to install::
---
- hosts: all
sudo: yes
gather_facts: no
connection: ssh
tasks:
- easy_install: name=pip
- pip: name={{ item }} state=present
with_items:
- pyzmq
- pyasn1
- PyCrypto
- python-keyczar
Fedora and EPEL also have Ansible RPM subpackages available for fireball-dependencies.
Also see the module documentation section.
Understanding Variable Precedence
`````````````````````````````````
You have already learned about inventory variables, 'vars', and 'vars_files'. In the
event the same variable name occurs in more than one place, what happens? There are really three tiers
of precedence, and within those tiers, some minor ordering rules that you probably won't even need to remember.
We'll explain them anyway though.
Variables that are set during the execution of the play have highest priority. This includes registered
variables and facts, which are discovered pieces of information about remote hosts.
Descending in priority are variables defined in the playbook. 'vars_files' as defined in the playbook are next up,
followed by variables as passed to ansible-playbook via --extra-vars (-e), then variables defined in the 'vars' section. These
should all be taken to be basically the same thing -- good places to define constants about what the play does to all hosts
in the play.
Finally, inventory variables have the least priority. Variables about hosts override those about groups.
If a variable is defined in multiple groups and one group is a child of the other, the child group variable
will override the variable set in the parent.
This makes the 'group_vars/all' file the best place to define a default value you wish to override in another
group, or even in a playbook. For example, your organization might set a default ntp server in group_vars/all
and then override it based on a group based on a geographic region. However if you type 'ntpserver: asdf.example.com'
in a vars section of a playbook, you know from reading the playbook that THAT specific value is definitely the one
that is going to be used. You won't be fooled by some variable from inventory sneaking up on you.
So, in short, if you want something easy to remember: facts beat playbook definitions, and
playbook definitions beat inventory variables.
There's a little bit more if you are using roles -- roles fit in the "playbook definitions" category of scale. They are
trumped by facts, and still trump inventory variables. However, there's a bit of extra magic.
Variables passed as parameters to the role are accesible only within that role (and dependencies of that role). You can
almost think of them like programming functions or macros.
Variables loaded via the 'vars/' directory of a role are made available to all roles and tasks, which in older versions of Ansible
could be confusing in the case of a reused variable name. In Ansible 1.3 and later, however, vars/ directories are guaranteed to be scoped to the current role, just like roles parameters. They are still available globally though, so if you want to set a variable like "ntp_server" in a common role, other roles can still make use of it. Thus they are just like "vars_files" construct that they emulate, but they have a bit more of a "Do What I Mean" semantic to them. They are smarter.
If there are role dependencies involved, dependent roles can set variables visible to the roles that require them, but
the requiring role is allowed to override those variables. For instance if a role "myapp" requires "apache", and
the value of "apache_port" in "apache" is 80, "myapp" could choose to set it to 8080. Thus you may think of this somewhat
like an inheritance system if you're a developer -- though it's not exactly -- and we don't require folks to think in programming terms to know how things work.
If you want, you can choose to prefix variable names with the name of your role and be extra sure of where
data sources are coming from, but this is optional. However it can be a nice thing to do in your templates as you immediately
know where the variable was defined.
Ultimately, the variable system may seem complex -- but it's really not. It's mostly a "Do What I Mean" kind of system, though knowing the details may help you if you get stuck or are trying to do something advanced. Feel free to experiment!
Check Mode ("Dry Run") --check
```````````````````````````````
.. versionadded:: 1.1
When ansible-playbook is executed with --check it will not make any changes on remote systems. Instead, any module
instrumented to support 'check mode' (which contains the primary core modules, but it is not required that all modules do
this) will report what changes they would have made. Other modules that do not support check mode will also take no
action, but just will not report what changes they might have made.
Check mode is just a simulation, and if you have steps that use conditionals that depend on the results of prior commands,
it may be less useful for you. However it is great for one-node-at-time basic configuration management use cases.
Example::
ansible-playbook foo.yml --check
Running a task in check mode
````````````````````````````
.. versionadded:: 1.3
Sometimes you may want to have a task to be executed even in check
mode. To achieve this, use the `always_run` clause on the task. Its
value is a Jinja2 expression, just like the `when` clause. In simple
cases a boolean YAML value would be sufficient as a value.
Example::
tasks:
- name: this task is run even in check mode
command: /something/to/run --even-in-check-mode
always_run: yes
As a reminder, a task with a `when` clause evaluated to false, will
still be skipped even if it has a `always_run` clause evaluated to
true.
Showing Differences with --diff
```````````````````````````````
.. versionadded:: 1.1
The --diff option to ansible-playbook works great with --check (detailed above) but can also be used by itself. When this flag is supplied, if any templated files on the remote system are changed, and the ansible-playbook CLI will report back
the textual changes made to the file (or, if used with --check, the changes that would have been made). Since the diff
feature produces a large amount of output, it is best used when checking a single host at a time, like so::
As discussed in the playbooks chapter, Ansible facts are a way of getting data about remote systems for use in playbook variables.
Usually these are discovered automatically by the 'setup' module in Ansible. Users can also write custom facts modules, as described
in the API guide. However, what if you want to have a simple way to provide system or user
provided data for use in Ansible variables, without writing a fact module? For instance, what if you want users to be able to control some aspect about how their systems are managed? "Facts.d" is one such mechanism.
If a remotely managed system has an "/etc/ansible/facts.d" directory, any files in this directory
ending in ".fact", can be JSON, INI, or executable files returning JSON, and these can supply local facts in Ansible.
For instance assume a /etc/ansible/facts.d/preferences.fact::
[general]
asdf=1
bar=2
This will produce a hash variable fact named "general" with 'asdf' and 'bar' as members.
To validate this, run the following::
ansible <hostname> -m setup -a "filter=ansible_local"
And you will see the following fact added::
"ansible_local": {
"preferences": {
"general": {
"asdf" : "1",
"bar" : "2"
}
}
}
And this data can be accessed in a template/playbook as::
{{ ansible_local.preferences.general.asdf }}
The local namespace prevents any user supplied fact from overriding system facts
or variables defined elsewhere in the playbook.
Style Points
````````````
Ansible playbooks are colorized. If you do not like this, set the ANSIBLE_NOCOLOR=1 environment variable.
Ansible playbooks also look more impressive with cowsay installed, and we encourage installing this package.
.. seealso::
:doc:`YAMLSyntax`
Learn about YAML syntax
:doc:`playbooks`
Review the basic playbook features
:doc:`bestpractices`
Various tips about playbooks in the real world
:doc:`modules`
Learn about available modules
:doc:`moduledev`
Learn how to extend Ansible by writing your own modules
Additionally, *group_names* is a list (array) of all the groups the current host is in. This can be used in templates using Jinja2 syntax to make template source files that vary based on the group membership (or role) of the host::
{% if 'webserver' in group_names %}
# some part of a configuration file that only applies to webservers
{% endif %}
*groups* is a list of all the groups (and hosts) in the inventory. This can be used to enumerate all hosts within a group.
For example::
{% for host in groups['app_servers'] %}
# something that applies to all app servers.
{% endfor %}
A frequently used idiom is walking a group to find all IP addresses in that group::
An example of this could include pointing a frontend proxy server to all of the app servers, setting up the correct firewall rules between servers, etc.
Just a few other 'magic' variables are available... There aren't many.
Additionally, *inventory_hostname* is the name of the hostname as configured in Ansible's inventory host file. This can
be useful for when you don't want to rely on the discovered hostname `ansible_hostname` or for other mysterious
reasons. If you have a long FQDN, *inventory_hostname_short* also contains the part up to the first
period, without the rest of the domain.
Don't worry about any of this unless you think you need it. You'll know when you do.
Also available, *inventory_dir* is the pathname of the directory holding Ansible's inventory host file, *inventory_file* is the pathname and the filename pointing to the Ansible's inventory host file.
Variable File Separation
````````````````````````
It's a great idea to keep your playbooks under source control, but
you may wish to make the playbook source public while keeping certain
important variables private. Similarly, sometimes you may just
want to keep certain information in different files, away from
the main playbook.
You can do this by using an external variables file, or files, just like this::
---
- hosts: all
remote_user: root
vars:
favcolor: blue
vars_files:
- /vars/external_vars.yml
tasks:
- name: this is just a placeholder
command: /bin/echo foo
This removes the risk of sharing sensitive data with others when
sharing your playbook source with them.
The contents of each variables file is a simple YAML dictionary, like this::
---
# in the above example, this would be vars/external_vars.yml
somevar: somevalue
password: magic
.. note::
It's also possible to keep per-host and per-group variables in very
similar files, this is covered in :ref:`patterns`.
Prompting For Sensitive Data
````````````````````````````
You may wish to prompt the user for certain input, and can
do so with the similarly named 'vars_prompt' section. This has uses
beyond security, for instance, you may use the same playbook for all
software releases and would prompt for a particular release version
in a push-script::
---
- hosts: all
remote_user: root
vars:
from: "camelot"
vars_prompt:
name: "what is your name?"
quest: "what is your quest?"
favcolor: "what is your favorite color?"
There are full examples of both of these items in the github examples/playbooks directory.
If you have a variable that changes infrequently, it might make sense to
provide a default value that can be overridden. This can be accomplished using
the default argument::
vars_prompt:
- name: "release_version"
prompt: "Product release version"
default: "1.0"
An alternative form of vars_prompt allows for hiding input from the user, and may later support
some other options, but otherwise works equivalently::
vars_prompt:
- name: "some_password"
prompt: "Enter password"
private: yes
- name: "release_version"
prompt: "Product release version"
private: no
If `Passlib <http://pythonhosted.org/passlib/>`_ is installed, vars_prompt can also crypt the
entered value so you can use it, for instance, with the user module to define a password::
vars_prompt:
- name: "my_password2"
prompt: "Enter password2"
private: yes
encrypt: "md5_crypt"
confirm: yes
salt_size: 7
You can use any crypt scheme supported by 'Passlib':
As with the case of 'with_items' above, you can use previously defined variables. Just specify the variable's name without templating it with '{{ }}'::
- name: here, 'users' contains the above list of employees
``with_password`` and associated lookup macro generate a random plaintext password and store it in
a file at a given filepath. Support for crypted save modes (as with vars_prompt) is pending. If the
file exists previously, it will retrieve its contents, behaving just like with_file. Usage of variables like "{{ inventory_hostname }}" in the filepath can be used to set
up random passwords per host (what simplifies password management in 'host_vars' variables).
Generated passwords contain a random mix of upper and lowercase ASCII letters, the
numbers 0-9 and punctuation (". , : - _"). The default length of a generated password is 30 characters.
This length can be changed by passing an extra parameter::
It is quite possible that you may need to get package updates through a proxy, or even get some package
updates through a proxy and access other packages not through a proxy. Ansible makes it easy for you
to configure your environment by using the 'environment' keyword. Here is an example::
- hosts: all
remote_user: root
tasks:
- apt: name=cobbler state=installed
environment:
http_proxy: http://proxy.example.com:8080
The environment can also be stored in a variable, and accessed like so::
- hosts: all
remote_user: root
# here we make a variable named "env" that is a dictionary
vars:
proxy_env:
http_proxy: http://proxy.example.com:8080
tasks:
- apt: name=cobbler state=installed
environment: "{{ proxy_env }}"
While just proxy settings were shown above, any number of settings can be supplied. The most logical place
to define an environment hash might be a group_vars file, like so::
---
# file: group_vars/boston
ntp_server: ntp.bos.example.com
backup: bak.bos.example.com
proxy_env:
http_proxy: http://proxy.bos.example.com:8080
https_proxy: http://proxy.bos.example.com:8080
Getting values from files
`````````````````````````
.. versionadded:: 0.8
Sometimes you'll want to include the content of a file directly into a playbook. You can do so using a macro.
This syntax will remain in future versions, though we will also will provide ways to do this via lookup plugins (see "More Loops") as well. What follows
is an example using the authorized_key module, which requires the actual text of the SSH key as a parameter::
Sometimes a configuration file you want to copy, or a template you will use may depend on a variable.
The following construct selects the first available file appropriate for the variables of a given host, which is often much cleaner than putting a lot of if conditionals in a template.
The following example shows how to template out a configuration file that was very different between, say, CentOS and Debian::
first_available_file is only available to the copy and template modules.
Asynchronous Actions and Polling
````````````````````````````````
By default tasks in playbooks block, meaning the connections stay open
until the task is done on each node. If executing playbooks with
a small parallelism value (aka ``--forks``), you may wish that long
running operations can go faster. The easiest way to do this is
to kick them off all at once and then poll until they are done.
You will also want to use asynchronous mode on very long running
operations that might be subject to timeout.
To launch a task asynchronously, specify its maximum runtime
and how frequently you would like to poll for status. The default
poll value is 10 seconds if you do not specify a value for `poll`::
---
- hosts: all
remote_user: root
tasks:
- name: simulate long running op (15 sec), wait for up to 45, poll every 5
command: /bin/sleep 15
async: 45
poll: 5
.. note::
There is no default for the async time limit. If you leave off the
'async' keyword, the task runs synchronously, which is Ansible's
default.
Alternatively, if you do not need to wait on the task to complete, you may
"fire and forget" by specifying a poll value of 0::
---
- hosts: all
remote_user: root
tasks:
- name: simulate long running op, allow to run for 45, fire and forget
command: /bin/sleep 15
async: 45
poll: 0
.. note::
You shouldn't "fire and forget" with operations that require
exclusive locks, such as yum transactions, if you expect to run other
commands later in the playbook against those same resources.
.. note::
Using a higher value for ``--forks`` will result in kicking off asynchronous
tasks even faster. This also increases the efficiency of polling.
Local Playbooks
```````````````
It may be useful to use a playbook locally, rather than by connecting over SSH. This can be useful
for assuring the configuration of a system by putting a playbook on a crontab. This may also be used
to run a playbook inside a OS installer, such as an Anaconda kickstart.
To run an entire playbook locally, just set the "hosts:" line to "hosts:127.0.0.1" and then run the playbook like so::
ansible-playbook playbook.yml --connection=local
Alternatively, a local connection can be used in a single playbook play, even if other plays in the playbook
use the default remote connection type::
hosts: 127.0.0.1
connection: local
Turning Off Facts
`````````````````
If you know you don't need any fact data about your hosts, and know everything about your systems centrally, you
can turn off fact gathering. This has advantages in scaling Ansible in push mode with very large numbers of
systems, mainly, or if you are using Ansible on experimental platforms. In any play, just do this::
- hosts: whatever
gather_facts: no
Pull-Mode Playbooks
```````````````````
The use of playbooks in local mode (above) is made extremely powerful with the addition of `ansible-pull`.
A script for setting up ansible-pull is provided in the examples/playbooks directory of the source
checkout.
The basic idea is to use Ansible to set up a remote copy of Ansible on each managed node, each set to run via
cron and update playbook source via git. This inverts the default push architecture of Ansible into a pull
architecture, which has near-limitless scaling potential. The setup playbook can be tuned to change
the cron frequency, logging locations, and parameters to ansible-pull.
This is useful both for extreme scale-out as well as periodic remediation. Usage of the 'fetch' module to retrieve
logs from ansible-pull runs would be an excellent way to gather and analyze remote logs from ansible-pull.
Register Variables
``````````````````
.. versionadded:: 0.7
Often in a playbook it may be useful to store the result of a given command in a variable and access
it later. Use of the command module in this way can in many ways eliminate the need to write site specific facts, for
instance, you could test for the existence of a particular program.
The 'register' keyword decides what variable to save a result in. The resulting variables can be used in templates, action lines, or *when* statements. It looks like this (in an obviously trivial example)::
- name: test play
hosts: all
tasks:
- shell: cat /etc/motd
register: motd_contents
- shell: echo "motd contains the word hi"
when: motd_contents.stdout.find('hi') != -1
As shown previously, the registered variable's string contents are accessible with the 'stdout' value.
The registered result can be used in the "with_items" of a task if it is converted into
a list (or already is a list) as shown below. "stdout_lines" is already available on the object as
well though you could also call "home_dirs.stdout.split()" if you wanted, and could split by other
fields::
- name: registered variable usage as a with_items list
A common pattern is to use a local action to call 'rsync' to recursively copy files to the managed servers.
Here is an example::
---
# ...
tasks:
- name: recursively copy files from management server to target
local_action: command rsync -a /path/to/files {{ inventory_hostname }}:/path/to/target/
Note that you must have passphrase-less SSH keys or an ssh-agent configured for this to work, otherwise rsync
will need to ask for a passphrase.
Accelerated Mode
````````````````
.. versionadded:: 1.3
While SSH using the ControlPersist feature is quite fast and scalable, there is a certain amount of overhead involved in
creating connections. This can become something of a bottleneck when the number of hosts grows into the hundreds or
thousands. To help overcome this, Ansible offers an accelerated connection option. Accelerated mode can be anywhere from
2-6x faster than SSH with ControlPersist enabled, and 10x faster than paramiko.
Accelerated mode works by launching a temporary daemon over SSH. Once the daemon is running, Ansible will connect directly
to it via a raw socket connection. Ansible secures this communication by using a temporary AES key that is uploaded during
the SSH connection (this key is different for every host, and is also regenerated every time the daemon is started). By default,
Ansible will use port 5099 for the accelerated connection, though this is configurable. Once running, the daemon will accept
connections for 30 minutes, after which time it will terminate itself and need to be restarted over SSH.
Accelerated mode offers several improvments over the original fireball mode:
* No bootstrapping is required, only a single line needs to be added to each play you wish to run in accelerated mode.
* Support for sudo commands (see below for more details and caveats).
* Fewer requirements! ZeroMQ is no longer required, nor are there any special packages beyond python-keyczar.
In order to use accelerated mode, simply add `accelerate: true` to your play::
---
- hosts: all
accelerate: true
tasks:
- name: some task
command: echo {{ item }}
with_items:
- foo
- bar
- baz
If you wish to change the port Ansible will use for the accelerated connection, just add the `accelerated_port` option::
---
- hosts: all
accelerate: true
# default port is 5099
accelerate_port: 10000
The `accelerate_port` option can also be specified in the environment variable ACCELERATE_PORT, or in your `ansible.cfg` configuration::
[accelerate]
accelerate_port = 5099
As noted above, accelerated mode also supports running tasks via sudo, however there are two important caveats:
* You must remove requiretty from your sudoers options.
* Prompting for the sudo password is not yet supported, so the NOPASSWD option is required for commands.
Fireball Mode
`````````````
.. versionadded:: 0.8 (deprecated as of 1.3)
.. note::
The following section has been deprecated as of Ansible 1.3 in favor of the accelerated mode described above. This
documentation is here for users who may still be using the original fireball connection method only, and should not
be used for any new deployments.
Ansible's core connection types of 'local', 'paramiko', and 'ssh' are augmented in version 0.8 and later by a new extra-fast
connection type called 'fireball'. It can only be used with playbooks and does require some additional setup
outside the lines of Ansible's normal "no bootstrapping" philosophy. You are not required to use fireball mode
to use Ansible, though some users may appreciate it.
Fireball mode works by launching a temporary 0mq daemon from SSH that by default lives for only 30 minutes before
shutting off. Fireball mode, once running, uses temporary AES keys to encrypt a session, and requires direct
communication to given nodes on the configured port. The default is 5099. The fireball daemon runs as any user you
set it down as. So it can run as you, root, or so on. If multiple users are running Ansible as the same batch of hosts,
take care to use unique ports.
Fireball mode is roughly 10 times faster than paramiko for communicating with nodes and may be a good option
if you have a large number of hosts::
---
# set up the fireball transport
- hosts: all
gather_facts: no
connection: ssh # or paramiko
sudo: yes
tasks:
- action: fireball
# these operations will occur over the fireball transport
- hosts: all
connection: fireball
tasks:
- shell: echo "Hello {{ item }}"
with_items:
- one
- two
In order to use fireball mode, certain dependencies must be installed on both ends. You can use this playbook as a basis for initial bootstrapping on
any platform. You will also need gcc and zeromq-devel installed from your package manager, which you can of course also get Ansible to install::
---
- hosts: all
sudo: yes
gather_facts: no
connection: ssh
tasks:
- easy_install: name=pip
- pip: name={{ item }} state=present
with_items:
- pyzmq
- pyasn1
- PyCrypto
- python-keyczar
Fedora and EPEL also have Ansible RPM subpackages available for fireball-dependencies.
Also see the module documentation section.
Understanding Variable Precedence
`````````````````````````````````
You have already learned about inventory variables, 'vars', and 'vars_files'. In the
event the same variable name occurs in more than one place, what happens? There are really three tiers
of precedence, and within those tiers, some minor ordering rules that you probably won't even need to remember.
We'll explain them anyway though.
Variables that are set during the execution of the play have highest priority. This includes registered
variables and facts, which are discovered pieces of information about remote hosts.
Descending in priority are variables defined in the playbook. 'vars_files' as defined in the playbook are next up,
followed by variables as passed to ansible-playbook via --extra-vars (-e), then variables defined in the 'vars' section. These
should all be taken to be basically the same thing -- good places to define constants about what the play does to all hosts
in the play.
Finally, inventory variables have the least priority. Variables about hosts override those about groups.
If a variable is defined in multiple groups and one group is a child of the other, the child group variable
will override the variable set in the parent.
This makes the 'group_vars/all' file the best place to define a default value you wish to override in another
group, or even in a playbook. For example, your organization might set a default ntp server in group_vars/all
and then override it based on a group based on a geographic region. However if you type 'ntpserver: asdf.example.com'
in a vars section of a playbook, you know from reading the playbook that THAT specific value is definitely the one
that is going to be used. You won't be fooled by some variable from inventory sneaking up on you.
So, in short, if you want something easy to remember: facts beat playbook definitions, and
playbook definitions beat inventory variables.
There's a little bit more if you are using roles -- roles fit in the "playbook definitions" category of scale. They are
trumped by facts, and still trump inventory variables. However, there's a bit of extra magic.
Variables passed as parameters to the role are accesible only within that role (and dependencies of that role). You can
almost think of them like programming functions or macros.
Variables loaded via the 'vars/' directory of a role are made available to all roles and tasks, which in older versions of Ansible
could be confusing in the case of a reused variable name. In Ansible 1.3 and later, however, vars/ directories are guaranteed to be scoped to the current role, just like roles parameters. They are still available globally though, so if you want to set a variable like "ntp_server" in a common role, other roles can still make use of it. Thus they are just like "vars_files" construct that they emulate, but they have a bit more of a "Do What I Mean" semantic to them. They are smarter.
If there are role dependencies involved, dependent roles can set variables visible to the roles that require them, but
the requiring role is allowed to override those variables. For instance if a role "myapp" requires "apache", and
the value of "apache_port" in "apache" is 80, "myapp" could choose to set it to 8080. Thus you may think of this somewhat
like an inheritance system if you're a developer -- though it's not exactly -- and we don't require folks to think in programming terms to know how things work.
If you want, you can choose to prefix variable names with the name of your role and be extra sure of where
data sources are coming from, but this is optional. However it can be a nice thing to do in your templates as you immediately
know where the variable was defined.
Ultimately, the variable system may seem complex -- but it's really not. It's mostly a "Do What I Mean" kind of system, though knowing the details may help you if you get stuck or are trying to do something advanced. Feel free to experiment!
Check Mode ("Dry Run") --check
```````````````````````````````
.. versionadded:: 1.1
When ansible-playbook is executed with --check it will not make any changes on remote systems. Instead, any module
instrumented to support 'check mode' (which contains the primary core modules, but it is not required that all modules do
this) will report what changes they would have made. Other modules that do not support check mode will also take no
action, but just will not report what changes they might have made.
Check mode is just a simulation, and if you have steps that use conditionals that depend on the results of prior commands,
it may be less useful for you. However it is great for one-node-at-time basic configuration management use cases.
Example::
ansible-playbook foo.yml --check
Running a task in check mode
````````````````````````````
.. versionadded:: 1.3
Sometimes you may want to have a task to be executed even in check
mode. To achieve this, use the `always_run` clause on the task. Its
value is a Jinja2 expression, just like the `when` clause. In simple
cases a boolean YAML value would be sufficient as a value.
Example::
tasks:
- name: this task is run even in check mode
command: /something/to/run --even-in-check-mode
always_run: yes
As a reminder, a task with a `when` clause evaluated to false, will
still be skipped even if it has a `always_run` clause evaluated to
true.
Showing Differences with --diff
```````````````````````````````
.. versionadded:: 1.1
The --diff option to ansible-playbook works great with --check (detailed above) but can also be used by itself. When this flag is supplied, if any templated files on the remote system are changed, and the ansible-playbook CLI will report back
the textual changes made to the file (or, if used with --check, the changes that would have been made). Since the diff
feature produces a large amount of output, it is best used when checking a single host at a time, like so::
As discussed in the playbooks chapter, Ansible facts are a way of getting data about remote systems for use in playbook variables.
Usually these are discovered automatically by the 'setup' module in Ansible. Users can also write custom facts modules, as described
in the API guide. However, what if you want to have a simple way to provide system or user
provided data for use in Ansible variables, without writing a fact module? For instance, what if you want users to be able to control some aspect about how their systems are managed? "Facts.d" is one such mechanism.
If a remotely managed system has an "/etc/ansible/facts.d" directory, any files in this directory
ending in ".fact", can be JSON, INI, or executable files returning JSON, and these can supply local facts in Ansible.
For instance assume a /etc/ansible/facts.d/preferences.fact::
[general]
asdf=1
bar=2
This will produce a hash variable fact named "general" with 'asdf' and 'bar' as members.
To validate this, run the following::
ansible <hostname> -m setup -a "filter=ansible_local"
And you will see the following fact added::
"ansible_local": {
"preferences": {
"general": {
"asdf" : "1",
"bar" : "2"
}
}
}
And this data can be accessed in a template/playbook as::
{{ ansible_local.preferences.general.asdf }}
The local namespace prevents any user supplied fact from overriding system facts
or variables defined elsewhere in the playbook.
Style Points
````````````
You may also apply takes to roles:
Ansible playbooks are colorized. If you do not like this, set the ANSIBLE_NOCOLOR=1 environment variable.
