Updating docs, added extensive info on playbook async + external vars_files for keeping variables outside of

the playbook file.
This commit is contained in:
Michael DeHaan 2012-03-12 23:12:21 -04:00
parent cf4ac3b6b2
commit 37adadcc50
9 changed files with 447 additions and 160 deletions

View file

@ -320,7 +320,7 @@ Stop by the mailing list if you have ideas.</p>
example is for configuration management where you
are starting from a clean OS with no extra software installed, adopting systems
that are already deployed.</p>
<p>Ansible is also great for running ad-hoc tasks across a wide variety of Linux, Unix, and <a href="#id1"><span class="problematic" id="id2">*</span></a>BSDs.
<p>Ansible is also great for running ad-hoc tasks across a wide variety of Linux, Unix, and BSDs.
Because it just uses the basic tools available on the system, it is exceptionally cross platform
without needing to install management packages on each node.</p>
<p>It also excels for writing distributed

View file

@ -279,14 +279,19 @@ you with questions about Ansible.</p>
<li class="toctree-l1"><a class="reference internal" href="playbooks.html">Playbooks</a><ul>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#playbook-example">Playbook Example</a></li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#hosts-line">Hosts line</a></li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#user-line">User line</a></li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#vars-section">Vars section</a></li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#tasks-list">Tasks list</a></li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#task-name-and-action">Task name and action</a></li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#notify-statements">Notify statements</a></li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#handlers">Handlers</a></li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#includes">Includes</a></li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#using-includes-to-assign-classes-of-systems">Using Includes To Assign Classes of Systems</a></li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#asynchronous-actions-and-polling">Asynchronous Actions and Polling</a></li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#power-tricks">Power Tricks</a><ul>
<li class="toctree-l3"><a class="reference internal" href="playbooks.html#external-variables-and-sensitive-data">External Variables And Sensitive Data</a></li>
<li class="toctree-l3"><a class="reference internal" href="playbooks.html#include-files-and-reuse">Include Files And Reuse</a></li>
<li class="toctree-l3"><a class="reference internal" href="playbooks.html#using-includes-to-assign-classes-of-systems">Using Includes To Assign Classes of Systems</a></li>
<li class="toctree-l3"><a class="reference internal" href="playbooks.html#asynchronous-actions-and-polling">Asynchronous Actions and Polling</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="playbooks.html#executing-a-playbook">Executing A Playbook</a></li>
</ul>
</li>

View file

@ -1,6 +1,6 @@
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>ansible-playbook</title><link rel="stylesheet" href="./docbook-xsl.css" type="text/css" /><meta name="generator" content="DocBook XSL Stylesheets V1.75.2" /></head><body><div xml:lang="en" class="refentry" title="ansible-playbook" lang="en"><a id="id446525"></a><div class="titlepage"></div><div class="refnamediv"><h2>Name</h2><p>ansible-playbook — run an ansible playbook</p></div><div class="refsynopsisdiv" title="Synopsis"><a id="_synopsis"></a><h2>Synopsis</h2><p>ansible-playbook &lt;filename.yml&gt; … [options]</p></div><div class="refsect1" title="DESCRIPTION"><a id="_description"></a><h2>DESCRIPTION</h2><p><span class="strong"><strong>Ansible playbooks</strong></span> are a configuration and multinode deployment system. Ansible-playbook is the tool
<html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>ansible-playbook</title><link rel="stylesheet" href="./docbook-xsl.css" type="text/css" /><meta name="generator" content="DocBook XSL Stylesheets V1.75.2" /></head><body><div xml:lang="en" class="refentry" title="ansible-playbook" lang="en"><a id="id320402"></a><div class="titlepage"></div><div class="refnamediv"><h2>Name</h2><p>ansible-playbook — run an ansible playbook</p></div><div class="refsynopsisdiv" title="Synopsis"><a id="_synopsis"></a><h2>Synopsis</h2><p>ansible-playbook &lt;filename.yml&gt; … [options]</p></div><div class="refsect1" title="DESCRIPTION"><a id="_description"></a><h2>DESCRIPTION</h2><p><span class="strong"><strong>Ansible playbooks</strong></span> are a configuration and multinode deployment system. Ansible-playbook is the tool
used to run them. See the project home page (link below) for more information.</p></div><div class="refsect1" title="ARGUMENTS"><a id="_arguments"></a><h2>ARGUMENTS</h2><div class="variablelist"><dl><dt><span class="term">
<span class="strong"><strong>filename.yml</strong></span>
</span></dt><dd>

View file

@ -1,6 +1,6 @@
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>ansible</title><link rel="stylesheet" href="./docbook-xsl.css" type="text/css" /><meta name="generator" content="DocBook XSL Stylesheets V1.75.2" /></head><body><div xml:lang="en" class="refentry" title="ansible" lang="en"><a id="id302038"></a><div class="titlepage"></div><div class="refnamediv"><h2>Name</h2><p>ansible — run a command somewhere else</p></div><div class="refsynopsisdiv" title="Synopsis"><a id="_synopsis"></a><h2>Synopsis</h2><p>ansible &lt;host-pattern&gt; [-f forks] [-m module_name] [-a args]</p></div><div class="refsect1" title="DESCRIPTION"><a id="_description"></a><h2>DESCRIPTION</h2><p><span class="strong"><strong>Ansible</strong></span> is an extra-simple tool/framework/API for doing 'remote things' over
<html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /><title>ansible</title><link rel="stylesheet" href="./docbook-xsl.css" type="text/css" /><meta name="generator" content="DocBook XSL Stylesheets V1.75.2" /></head><body><div xml:lang="en" class="refentry" title="ansible" lang="en"><a id="id570190"></a><div class="titlepage"></div><div class="refnamediv"><h2>Name</h2><p>ansible — run a command somewhere else</p></div><div class="refsynopsisdiv" title="Synopsis"><a id="_synopsis"></a><h2>Synopsis</h2><p>ansible &lt;host-pattern&gt; [-f forks] [-m module_name] [-a args]</p></div><div class="refsect1" title="DESCRIPTION"><a id="_description"></a><h2>DESCRIPTION</h2><p><span class="strong"><strong>Ansible</strong></span> is an extra-simple tool/framework/API for doing 'remote things' over
SSH.</p></div><div class="refsect1" title="ARGUMENTS"><a id="_arguments"></a><h2>ARGUMENTS</h2><div class="variablelist"><dl><dt><span class="term">
<span class="strong"><strong>host-pattern</strong></span>
</span></dt><dd>

View file

@ -333,7 +333,7 @@ be a relative or absolute path.</li>
</ul>
<p><em>state</em>:</p>
<ul class="simple">
<li>Can be either &#8216;installed&#8217; or &#8216;removed&#8217;</li>
<li>Can be either &#8216;installed&#8217;, &#8216;latest&#8217;, or &#8216;removed&#8217;</li>
</ul>
<p><em>list</em>:</p>
<ul class="simple">

View file

@ -130,14 +130,19 @@ s.parentNode.insertBefore(ga, s);
<li><a class="reference internal" href="#">Playbooks</a><ul>
<li><a class="reference internal" href="#playbook-example">Playbook Example</a></li>
<li><a class="reference internal" href="#hosts-line">Hosts line</a></li>
<li><a class="reference internal" href="#user-line">User line</a></li>
<li><a class="reference internal" href="#vars-section">Vars section</a></li>
<li><a class="reference internal" href="#tasks-list">Tasks list</a></li>
<li><a class="reference internal" href="#task-name-and-action">Task name and action</a></li>
<li><a class="reference internal" href="#notify-statements">Notify statements</a></li>
<li><a class="reference internal" href="#handlers">Handlers</a></li>
<li><a class="reference internal" href="#includes">Includes</a></li>
<li><a class="reference internal" href="#power-tricks">Power Tricks</a><ul>
<li><a class="reference internal" href="#external-variables-and-sensitive-data">External Variables And Sensitive Data</a></li>
<li><a class="reference internal" href="#include-files-and-reuse">Include Files And Reuse</a></li>
<li><a class="reference internal" href="#using-includes-to-assign-classes-of-systems">Using Includes To Assign Classes of Systems</a></li>
<li><a class="reference internal" href="#asynchronous-actions-and-polling">Asynchronous Actions and Polling</a></li>
</ul>
</li>
<li><a class="reference internal" href="#executing-a-playbook">Executing A Playbook</a></li>
</ul>
</li>
@ -196,18 +201,22 @@ particularly awesome.</p>
<p>They are the basis for a really simple configuration management and
multi-machine deployment system, unlike any that already exist, and
one that is very well suited to deploying complex applications.</p>
<p>Playbooks can declare configurations, or they can automate steps of
a manual ordered process. They can launch tasks synchronously or asynchronously.</p>
<p>While you might run the main /usr/bin/ansible program for ad-hoc
tasks, playbooks are more likely to be kept in source control and used
to push out your configuration or assure the configurations of your
remote systems are in spec.</p>
<p>Let&#8217;s dive in and see how they work.</p>
<div class="section" id="playbook-example">
<h2>Playbook Example<a class="headerlink" href="#playbook-example" title="Permalink to this headline"></a></h2>
<p>Playbooks are expressed in YAML format and have a minimum of syntax.
Each playbook is composed of one or more &#8216;plays&#8217; in a list. By
composing a playbook of multiple &#8216;plays&#8217;, it is possible to
Each playbook is composed of one or more &#8216;plays&#8217; in a list.</p>
<p>By composing a playbook of multiple &#8216;plays&#8217;, it is possible to
orchestrate multi-machine deployments, running certain steps on all
machines in the webservers group, then certain steps on the database
server group, then more commands back on the webservers group, etc:</p>
server group, then more commands back on the webservers group, etc.</p>
<p>For starters, here&#8217;s a playbook that contains just one play.:</p>
<div class="highlight-python"><pre>---
- hosts: webservers
vars:
@ -216,8 +225,8 @@ server group, then more commands back on the webservers group, etc:</p>
user: root
tasks:
- include: base.yml somevar=3 othervar=4
- name: ensure apache is installed
action: yum pkg=httpd state=installed
- name: ensure apache is at the latest version
action: yum pkg=httpd state=latest
- name: write the apache config file
action: template src=/srv/httpd.j2 dest=/etc/httpd.conf
notify:
@ -227,19 +236,28 @@ server group, then more commands back on the webservers group, etc:</p>
handlers:
- include: handlers.yml</pre>
</div>
<p>Below, we&#8217;ll break down what the various features of the playbook language are.</p>
</div>
<div class="section" id="hosts-line">
<h2>Hosts line<a class="headerlink" href="#hosts-line" title="Permalink to this headline"></a></h2>
<p>The hosts line is a list of one or more groups or host patterns,
<p>The <cite>hosts</cite> line is a list of one or more groups or host patterns,
separated by colons, as described in the <a class="reference internal" href="patterns.html#patterns"><em>The Inventory File, Patterns, and Groups</em></a>
documentation. This is just like the first parameter to
<cite>/usr/bin/ansible</cite>.</p>
<p>Each play gets to designate it&#8217;s own choice of patterns.</p>
</div>
<div class="section" id="user-line">
<h2>User line<a class="headerlink" href="#user-line" title="Permalink to this headline"></a></h2>
<p>Playbook steps on the remote system can be executed as any user. The default is root,
but you can specify others. Sudo support is pending.:</p>
<div class="highlight-python"><pre>user: mdehaan</pre>
</div>
</div>
<div class="section" id="vars-section">
<h2>Vars section<a class="headerlink" href="#vars-section" title="Permalink to this headline"></a></h2>
<p>A list of variables and values that can be used in the plays. These
can be used in templates or &#8216;action&#8217; lines and are dereferenced using
<cite>jinja2</cite> syntax like this:</p>
<p>The <cite>vars&#8217; section contains a list of variables and values that can be used in the plays. These
can be used in templates or tasks and are dereferenced using
`jinja2</cite> syntax like this:</p>
<div class="highlight-python"><pre>{{ varname }}</pre>
</div>
<p>Further, if there are discovered variables about the system (say, if
@ -249,7 +267,7 @@ variables. Facter variables are prefixed with <tt class="docutils literal"><spa
variables are prefixed with <tt class="docutils literal"><span class="pre">ohai_</span></tt>. So for instance, if I wanted
to write the hostname into the /etc/motd file, I could say:</p>
<div class="highlight-python"><pre>- name: write the motd
- action: template src=/srv/templates/motd.j2 dest=/etc/motd</pre>
action: template src=/srv/templates/motd.j2 dest=/etc/motd</pre>
</div>
<p>And in /srv/templates/motd.j2:</p>
<div class="highlight-python"><pre>You are logged into {{ facter_hostname }}</pre>
@ -259,41 +277,60 @@ to write the hostname into the /etc/motd file, I could say:</p>
<div class="section" id="tasks-list">
<h2>Tasks list<a class="headerlink" href="#tasks-list" title="Permalink to this headline"></a></h2>
<p>Each play contains a list of tasks. Tasks are executed in order, one
at a time, against all machines matched by the playbooks host pattern,
at a time, against all machines matched by the host pattern,
before moving on to the next task.</p>
<p>Hosts with failed tasks are taken out of the rotation for the entire
playbook. If things fail, simply correct the playbook file and rerun.</p>
<p>Modules other than command are idempotent, meaning if you run them
<p>Modules other than <cite>command</cite> are &#8216;idempotent&#8217;, meaning if you run them
again, they will make the changes they are told to make to bring the
system to the desired state.</p>
system to the desired state. This makes it very safe to rerun
the same playbook multiple times. They won&#8217;t change things
unless they have to change things. Command will actually rerun the
same command again, which is totally ok if the command is something
like &#8216;chmod&#8217; or &#8216;setsebool&#8217;, etc.</p>
</div>
<div class="section" id="task-name-and-action">
<h2>Task name and action<a class="headerlink" href="#task-name-and-action" title="Permalink to this headline"></a></h2>
<p>Every task must have a name, which is included in the output from
running the playbook.</p>
<p>The action line is the name of an ansible module followed by
parameters. Usually these are expressed in <tt class="docutils literal"><span class="pre">key=value</span></tt> form, except
for the command module, which looks just like a Linux/Unix command
line. See the module documentation for more info.</p>
<p>Variables, as mentioned above, can be used in action lines. So if,
hypothetically, you wanted to make a directory on each system named
after the hostname ... yeah, that&#8217;s I know silly ... you could do it
like so:</p>
<div class="highlight-python"><pre>- name: make a directory
- action: mkdir /tmp/{{ facter_hostname }}</pre>
parameters in key=value form:</p>
<div class="highlight-python"><pre>- name: make sure apache is running
action: service name=httpd state=running</pre>
</div>
<p>The command module is the one module that just takes a list
of arguments, and doesn&#8217;t use the key=value form. Simple:</p>
<div class="highlight-python"><pre>- name: disable selinux
action: command /sbin/setenforce 0</pre>
</div>
<p>Variables can be used in action lines. Suppose you defined
a variable called &#8216;vhost&#8217; in the &#8216;vars&#8217; section, you could do this:</p>
<div class="highlight-python"><pre>- name: make a directory
action: template src=somefile.j2 dest=/etc/httpd/conf.d/{{ vhost }}</pre>
</div>
<p>Those same variables are usable in templates, which we&#8217;ll get to later.</p>
</div>
<div class="section" id="notify-statements">
<h2>Notify statements<a class="headerlink" href="#notify-statements" title="Permalink to this headline"></a></h2>
<p>Nearly all modules are written to be &#8216;idempotent&#8217; and can signal when
they have affected a change on the remote system. If a notify
statement is used, the named handler will be run against each system
where a change was effected, but NOT on systems where no change
occurred. This happens after all of the tasks are run. For example,
if notifying Apache and potentially replacing lots of configuration
files, you could have Apache restart just once, at the end of a run.
If you need Apache restarted in the middle of a run, you could just
make a task for it, no harm done. Notifiers are optional.</p>
<p>As we&#8217;ve mentioned, nearly all modules are written to be &#8216;idempotent&#8217; and can signal when
they have affected a change on the remote system. Playbooks recognize this and
have a basic event system that can be used to respond to change.</p>
<p>These &#8216;notify&#8217; actions are triggered at the end of each &#8216;play&#8217; in a playbook, and
trigger only once each. For instance, multiple resources may indicate
that apache needs to be restarted, but apache will only be bounced once.</p>
<p>Here&#8217;s an example of restarting two services when the contents of a file
change, but only if the file changes:</p>
<div class="highlight-python"><pre>- name: template configuration file
action: template src=template.j2 dest=/etc/foo.conf
notify:
- restart memcached
- restart foo</pre>
</div>
<p>Next up, we&#8217;ll show what a handler looks like.</p>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">Notify handlers are always run in the order written.</p>
</div>
</div>
<div class="section" id="handlers">
<h2>Handlers<a class="headerlink" href="#handlers" title="Permalink to this headline"></a></h2>
@ -302,51 +339,96 @@ tasks, that are referenced by name. Handlers are what notifiers
notify. If nothing notifies a handler, it will not run. Regardless
of how many things notify a handler, it will run only once, after all
of the tasks complete in a particular play.</p>
<p>Here&#8217;s an example handlers section:</p>
<div class="highlight-python"><pre>handlers:
- name: restart apache
action: service name=apache state=restarted
- name: restart memcached
action: service name=memcached state=restarted</pre>
</div>
<div class="section" id="includes">
<h2>Includes<a class="headerlink" href="#includes" title="Permalink to this headline"></a></h2>
<p>Not all tasks have to be listed directly in the main file. An include
file can contain a list of tasks (in YAML) as well, optionally passing
extra variables into the file. Variables passed in can be deferenced
like this (assume a variable named &#8216;user&#8217;):</p>
<p>Handlers are best used to restart services and trigger reboots. You probably
won&#8217;t need them for much else.</p>
</div>
<div class="section" id="power-tricks">
<h2>Power Tricks<a class="headerlink" href="#power-tricks" title="Permalink to this headline"></a></h2>
<p>Now that you have the basics down, let&#8217;s learn some more advanced
things you can do with playbooks.</p>
<div class="section" id="external-variables-and-sensitive-data">
<h3>External Variables And Sensitive Data<a class="headerlink" href="#external-variables-and-sensitive-data" title="Permalink to this headline"></a></h3>
<p>It&#8217;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. You can do this by using an external
variables file, or files, just like this:</p>
<div class="highlight-python"><pre>---
- hosts: all
user: root
vars:
favcolor: blue
vars_files:
- /path/to/external_vars.yml
tasks:
- name: this is just a placeholder
action: command /bin/echo foo</pre>
</div>
<p>This removes the risk of sharing sensitive data with others when
sharing your playbook source with them.</p>
<p>The contents of each variables file is a simple YAML dictionary, like this:</p>
<div class="highlight-python"><pre>---
somevar: somevalue
password: magic</pre>
</div>
</div>
<div class="section" id="include-files-and-reuse">
<h3>Include Files And Reuse<a class="headerlink" href="#include-files-and-reuse" title="Permalink to this headline"></a></h3>
<p>Suppose you want to reuse lists of tasks between plays or playbooks. You can use
include files to do this.</p>
<p>An include file simply contains a list of tasks, like so:</p>
<div class="highlight-python"><pre>---
- name: placeholder foo
action: command /bin/foo
- name: placeholder bar
action: command /bin/bar</pre>
</div>
<p>Variables passed in can be deferenced too. Assume a variable named &#8216;user&#8217;. Using
<cite>jinja2</cite> syntax, anywhere in the included file, you can say:</p>
<div class="highlight-python"><pre>{{ user }}</pre>
</div>
<p>For instance, if deploying multiple wordpress instances, I could
contain all of my tasks in a wordpress.yml file, and use it like so:</p>
contain all of my wordpress tasks in a single wordpress.yml file, and use it like so:</p>
<div class="highlight-python"><pre>- tasks:
- include: wordpress.yml user=timmy
- include: wordpress.yml user=alice
- include: wordpress.yml user=bob</pre>
- include: wordpress.yml user=timmy
- include: wordpress.yml user=alice
- include: wordpress.yml user=bob</pre>
</div>
<p>In addition to the explicitly passed in parameters, all variables from
the vars section are also available.</p>
<p>The format of an included list of tasks or handlers looks just like a
flat list of tasks. Here is an example of what base.yml might look
like:</p>
<div class="highlight-python"><pre>---
- name: no selinux
action: command /usr/sbin/setenforce 0
- name: no iptables
action: service name=iptables state=stopped
- name: this is just to show variables work here, favcolor={{ favcolor }}
action: command /bin/true</pre>
</div>
<p>As you can see above, variables in include files work just like they
do in the main file. Including a variable in the name of a task is a
contrived example, you could also pass them to the action command line
or use them inside a template file.</p>
the vars section are also available for use here as well. Variables that bubble
up from tools like facter and ohai are not though &#8211; but they ARE available for use
inside &#8216;action&#8217; lines.</p>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">Note that include statements are only usable from the top level
playbook file. At this time, includes can not include other
<p class="last">Include statements are only usable from the top level
playbook file. This means includes can not include other
includes.</p>
</div>
<p>Includes can also be used in the &#8216;handlers&#8217; section, for instance, if you
want to define how to restart apache, you only have to do that once for all
of your playbooks. You might make a notifiers.yaml that looked like:</p>
<div class="highlight-python"><pre>----
- name: restart apache
action: service name=apache state=restarted</pre>
</div>
<p>And in your main playbook file, just include it like so, at the bottom
of a play:</p>
<div class="highlight-python"><pre>handlers:
- include: handlers.yml</pre>
</div>
<p>You can mix in includes along with your regular non-included tasks and handlers.</p>
</div>
<div class="section" id="using-includes-to-assign-classes-of-systems">
<h2>Using Includes To Assign Classes of Systems<a class="headerlink" href="#using-includes-to-assign-classes-of-systems" title="Permalink to this headline"></a></h2>
<p>Include files are best used to reuse logic between playbooks. You
<h3>Using Includes To Assign Classes of Systems<a class="headerlink" href="#using-includes-to-assign-classes-of-systems" title="Permalink to this headline"></a></h3>
<p>Include files are really powerful when used to reuse logic between playbooks. You
could imagine a playbook describing your entire infrastructure like
this:</p>
this, in a list of just a few plays:</p>
<div class="highlight-python"><pre>---
- hosts: atlanta-webservers
vars:
@ -366,22 +448,71 @@ this:</p>
- include: generic-handlers.yml</pre>
</div>
<p>There is one (or more) play defined for each group of systems, and
each play maps each group includes one or more &#8216;class definitions&#8217;
telling the systems what they are supposed to do or be.</p>
<p>Using a common handlers file could allow one task in &#8216;webservers&#8217; to
define &#8216;restart apache&#8217;, and it could be reused between multiple
plays.</p>
<p>Variables like &#8216;database&#8217; above can be used in templates referenced
from the configuration file to generate machine specific variables.</p>
each play maps each group to several includes. These includes represent
&#8216;class definitions&#8217;, telling the systems what they are supposed to do or be.</p>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">Playbooks do not always have to be declarative; you can do something
similar to model a push process for a multi-tier web application. This is
actually one of the things playbooks were invented to do.</p>
</div>
</div>
<div class="section" id="asynchronous-actions-and-polling">
<h2>Asynchronous Actions and Polling<a class="headerlink" href="#asynchronous-actions-and-polling" title="Permalink to this headline"></a></h2>
<p>(Information on this feature is pending)</p>
<h3>Asynchronous Actions and Polling<a class="headerlink" href="#asynchronous-actions-and-polling" title="Permalink to this headline"></a></h3>
<p>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 <cite>&#8211;forks</cite>), 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.</p>
<p>You will also want to use asynchronous mode on very long running
operations that might be subject to timeout.</p>
<p>To launch a task asynchronously, specify it&#8217;s 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 <cite>poll</cite>:</p>
<div class="highlight-python"><pre>---
- hosts: all
user: root
tasks:
- name: simulate long running op (15 sec), wait for up to 45, poll every 5
action: command /bin/sleep 15
async: 45
poll: 5</pre>
</div>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">There is no default for the async time limit. If you leave off the
&#8216;async&#8217; keyword, the task runs synchronously, which is Ansible&#8217;s
default.</p>
</div>
<p>Alternatively, if you do not need to wait on the task to complete, you may
&#8220;fire and forget&#8221; by specifying a poll value of 0:</p>
<div class="highlight-python"><pre>---
- hosts: all
user: root
tasks:
- name: simulate long running op (15 sec), wait for up to 45, poll every 5
action: command /bin/sleep 15
async: 45
poll: 0</pre>
</div>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">You shouldn&#8217;t &#8220;fire and forget&#8221; 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.</p>
</div>
<div class="admonition note">
<p class="first admonition-title">Note</p>
<p class="last">Using a higher value for <cite>&#8211;forks</cite> will result in kicking off asynchronous
tasks even faster. This also increases the efficiency of polling.</p>
</div>
</div>
</div>
<div class="section" id="executing-a-playbook">
<h2>Executing A Playbook<a class="headerlink" href="#executing-a-playbook" title="Permalink to this headline"></a></h2>
<p>To run a playbook:</p>
<div class="highlight-python"><pre>ansible-playbook playbook.yml</pre>
<p>Now that you&#8217;ve learned playbook syntax, how do you run a playbook? It&#8217;s simple.
Let&#8217;s run a playbook using a parallelism level of 10:</p>
<div class="highlight-python"><pre>ansible-playbook playbook.yml -f 10</pre>
</div>
</div>
</div>

View file

@ -173,7 +173,7 @@ example is for configuration management where you
are starting from a clean OS with no extra software installed, adopting systems
that are already deployed.
Ansible is also great for running ad-hoc tasks across a wide variety of Linux, Unix, and *BSDs.
Ansible is also great for running ad-hoc tasks across a wide variety of Linux, Unix, and BSDs.
Because it just uses the basic tools available on the system, it is exceptionally cross platform
without needing to install management packages on each node.

View file

@ -18,21 +18,29 @@ They are the basis for a really simple configuration management and
multi-machine deployment system, unlike any that already exist, and
one that is very well suited to deploying complex applications.
Playbooks can declare configurations, or they can automate steps of
a manual ordered process. They can launch tasks synchronously or asynchronously.
While you might run the main /usr/bin/ansible program for ad-hoc
tasks, playbooks are more likely to be kept in source control and used
to push out your configuration or assure the configurations of your
remote systems are in spec.
Let's dive in and see how they work.
Playbook Example
````````````````
Playbooks are expressed in YAML format and have a minimum of syntax.
Each playbook is composed of one or more 'plays' in a list. By
composing a playbook of multiple 'plays', it is possible to
Each playbook is composed of one or more 'plays' in a list.
By composing a playbook of multiple 'plays', it is possible to
orchestrate multi-machine deployments, running certain steps on all
machines in the webservers group, then certain steps on the database
server group, then more commands back on the webservers group, etc::
server group, then more commands back on the webservers group, etc.
For starters, here's a playbook that contains just one play.::
---
- hosts: webservers
@ -53,22 +61,36 @@ server group, then more commands back on the webservers group, etc::
handlers:
- include: handlers.yml
Hosts line
``````````
The hosts line is a list of one or more groups or host patterns,
Below, we'll break down what the various features of the playbook language are.
Hosts line
```````````
The `hosts` line is a list of one or more groups or host patterns,
separated by colons, as described in the :ref:`patterns`
documentation. This is just like the first parameter to
`/usr/bin/ansible`.
`/usr/bin/ansible`.
Each play gets to designate it's own choice of patterns.
User line
`````````
Playbook steps on the remote system can be executed as any user. The default is root,
but you can specify others. Sudo support is pending.::
user: mdehaan
Vars section
````````````
A list of variables and values that can be used in the plays. These
can be used in templates or 'action' lines and are dereferenced using
The `vars' section contains a list of variables and values that can be used in the plays. These
can be used in templates or tasks and are dereferenced using
`jinja2` syntax like this::
{{ varname }}
{{ varname }}
Further, if there are discovered variables about the system (say, if
facter or ohai were installed) these variables bubble up back into the
@ -78,7 +100,7 @@ variables are prefixed with ``ohai_``. So for instance, if I wanted
to write the hostname into the /etc/motd file, I could say::
- name: write the motd
- action: template src=/srv/templates/motd.j2 dest=/etc/motd
action: template src=/srv/templates/motd.j2 dest=/etc/motd
And in /srv/templates/motd.j2::
@ -90,15 +112,20 @@ Tasks list
``````````
Each play contains a list of tasks. Tasks are executed in order, one
at a time, against all machines matched by the playbooks host pattern,
at a time, against all machines matched by the host pattern,
before moving on to the next task.
Hosts with failed tasks are taken out of the rotation for the entire
playbook. If things fail, simply correct the playbook file and rerun.
Modules other than command are idempotent, meaning if you run them
Modules other than `command` are 'idempotent', meaning if you run them
again, they will make the changes they are told to make to bring the
system to the desired state.
system to the desired state. This makes it very safe to rerun
the same playbook multiple times. They won't change things
unless they have to change things. Command will actually rerun the
same command again, which is totally ok if the command is something
like 'chmod' or 'setsebool', etc.
Task name and action
`````````````````````
@ -107,30 +134,49 @@ Every task must have a name, which is included in the output from
running the playbook.
The action line is the name of an ansible module followed by
parameters. Usually these are expressed in ``key=value`` form, except
for the command module, which looks just like a Linux/Unix command
line. See the module documentation for more info.
parameters in key=value form::
Variables, as mentioned above, can be used in action lines. So if,
hypothetically, you wanted to make a directory on each system named
after the hostname ... yeah, that's I know silly ... you could do it
like so::
- name: make sure apache is running
action: service name=httpd state=running
The command module is the one module that just takes a list
of arguments, and doesn't use the key=value form. Simple::
- name: disable selinux
action: command /sbin/setenforce 0
Variables can be used in action lines. Suppose you defined
a variable called 'vhost' in the 'vars' section, you could do this::
- name: make a directory
- action: mkdir /tmp/{{ facter_hostname }}
action: template src=somefile.j2 dest=/etc/httpd/conf.d/{{ vhost }}
Those same variables are usable in templates, which we'll get to later.
Notify statements
`````````````````
Nearly all modules are written to be 'idempotent' and can signal when
they have affected a change on the remote system. If a notify
statement is used, the named handler will be run against each system
where a change was effected, but NOT on systems where no change
occurred. This happens after all of the tasks are run. For example,
if notifying Apache and potentially replacing lots of configuration
files, you could have Apache restart just once, at the end of a run.
If you need Apache restarted in the middle of a run, you could just
make a task for it, no harm done. Notifiers are optional.
As we've mentioned, nearly all modules are written to be 'idempotent' and can signal when
they have affected a change on the remote system. Playbooks recognize this and
have a basic event system that can be used to respond to change.
These 'notify' actions are triggered at the end of each 'play' in a playbook, and
trigger only once each. For instance, multiple resources may indicate
that apache needs to be restarted, but apache will only be bounced once.
Here's an example of restarting two services when the contents of a file
change, but only if the file changes::
- name: template configuration file
action: template src=template.j2 dest=/etc/foo.conf
notify:
- restart memcached
- restart foo
Next up, we'll show what a handler looks like.
.. note::
Notify handlers are always run in the order written.
Handlers
````````
@ -139,57 +185,116 @@ Handlers are lists of tasks, not really any different from regular
tasks, that are referenced by name. Handlers are what notifiers
notify. If nothing notifies a handler, it will not run. Regardless
of how many things notify a handler, it will run only once, after all
of the tasks complete in a particular play.
of the tasks complete in a particular play.
Includes
````````
Here's an example handlers section::
Not all tasks have to be listed directly in the main file. An include
file can contain a list of tasks (in YAML) as well, optionally passing
extra variables into the file. Variables passed in can be deferenced
like this (assume a variable named 'user')::
handlers:
- name: restart apache
action: service name=apache state=restarted
- name: restart memcached
action: service name=memcached state=restarted
Handlers are best used to restart services and trigger reboots. You probably
won't need them for much else.
Power Tricks
````````````
Now that you have the basics down, let's learn some more advanced
things you can do with playbooks.
External Variables And Sensitive Data
+++++++++++++++++++++++++++++++++++++
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. You can do this by using an external
variables file, or files, just like this::
---
- hosts: all
user: root
vars:
favcolor: blue
vars_files:
- /path/to/external_vars.yml
tasks:
- name: this is just a placeholder
action: 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::
---
somevar: somevalue
password: magic
Include Files And Reuse
+++++++++++++++++++++++
Suppose you want to reuse lists of tasks between plays or playbooks. You can use
include files to do this.
An include file simply contains a list of tasks, like so::
---
- name: placeholder foo
action: command /bin/foo
- name: placeholder bar
action: command /bin/bar
Variables passed in can be deferenced too. Assume a variable named 'user'. Using
`jinja2` syntax, anywhere in the included file, you can say::
{{ user }}
For instance, if deploying multiple wordpress instances, I could
contain all of my tasks in a wordpress.yml file, and use it like so::
contain all of my wordpress tasks in a single wordpress.yml file, and use it like so::
- tasks:
- include: wordpress.yml user=timmy
- include: wordpress.yml user=alice
- include: wordpress.yml user=bob
- include: wordpress.yml user=timmy
- include: wordpress.yml user=alice
- include: wordpress.yml user=bob
In addition to the explicitly passed in parameters, all variables from
the vars section are also available.
The format of an included list of tasks or handlers looks just like a
flat list of tasks. Here is an example of what base.yml might look
like::
---
- name: no selinux
action: command /usr/sbin/setenforce 0
- name: no iptables
action: service name=iptables state=stopped
- name: this is just to show variables work here, favcolor={{ favcolor }}
action: command /bin/true
As you can see above, variables in include files work just like they
do in the main file. Including a variable in the name of a task is a
contrived example, you could also pass them to the action command line
or use them inside a template file.
the vars section are also available for use here as well. Variables that bubble
up from tools like facter and ohai are not though -- but they ARE available for use
inside 'action' lines.
.. note::
Note that include statements are only usable from the top level
playbook file. At this time, includes can not include other
includes.
Include statements are only usable from the top level
playbook file. This means includes can not include other
includes.
Includes can also be used in the 'handlers' section, for instance, if you
want to define how to restart apache, you only have to do that once for all
of your playbooks. You might make a notifiers.yaml that looked like::
----
- name: restart apache
action: service name=apache state=restarted
And in your main playbook file, just include it like so, at the bottom
of a play::
handlers:
- include: handlers.yml
You can mix in includes along with your regular non-included tasks and handlers.
Using Includes To Assign Classes of Systems
```````````````````````````````````````````
+++++++++++++++++++++++++++++++++++++++++++
Include files are best used to reuse logic between playbooks. You
Include files are really powerful when used to reuse logic between playbooks. You
could imagine a playbook describing your entire infrastructure like
this::
this, in a list of just a few plays::
---
- hosts: atlanta-webservers
@ -210,26 +315,72 @@ this::
- include: generic-handlers.yml
There is one (or more) play defined for each group of systems, and
each play maps each group includes one or more 'class definitions'
telling the systems what they are supposed to do or be.
each play maps each group to several includes. These includes represent
'class definitions', telling the systems what they are supposed to do or be.
Using a common handlers file could allow one task in 'webservers' to
define 'restart apache', and it could be reused between multiple
plays.
.. note::
Playbooks do not always have to be declarative; you can do something
similar to model a push process for a multi-tier web application. This is
actually one of the things playbooks were invented to do.
Variables like 'database' above can be used in templates referenced
from the configuration file to generate machine specific variables.
Asynchronous Actions and Polling
````````````````````````````````
++++++++++++++++++++++++++++++++
(Information on this feature is pending)
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 it's 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
user: root
tasks:
- name: simulate long running op (15 sec), wait for up to 45, poll every 5
action: 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
user: root
tasks:
- name: simulate long running op (15 sec), wait for up to 45, poll every 5
action: 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.
Executing A Playbook
````````````````````
To run a playbook::
Now that you've learned playbook syntax, how do you run a playbook? It's simple.
Let's run a playbook using a parallelism level of 10::
ansible-playbook playbook.yml -f 10
ansible-playbook playbook.yml

File diff suppressed because one or more lines are too long