Ansible

Overview
Questions:
  • Why Ansible?

  • How and when to use Ansible?

  • How to write a role?

  • How to leverage community build roles?

Objectives:
  • Learn Ansible basics

  • Write a simple role

  • Install a role from Ansible Galaxy

Time estimation: 60 minutes
Supporting Materials:
Last modification: Oct 18, 2022
License: Tutorial Content is licensed under Creative Commons Attribution 4.0 International License The GTN Framework is licensed under MIT

Overview

In this tutorial we will briefly cover what Ansible is and how to understand what it does. This guide is not meant to make you an expert on Ansible, but perhaps give you enough that you can debug broken roles and modify them to suit your needs. Or maybe to contribute to the Galaxyproject Ansible roles.

This will be a very practical training with emphasis on looking at examples from modules and becoming self sufficient.

Agenda
  1. Overview
  2. What is Ansible?
    1. Inventory file
    2. Roles
    3. Modules and Tasks
    4. Playbooks
    5. Variables
    6. Vaults
  3. Your First Playbook and First Role
    1. A Basic Role
    2. Facts
    3. Templates
  4. Ansible Galaxy
    1. Choosing a Role
  5. Ansible Vault
  6. Other Stuff
    1. With Items
    2. When Changed
    3. Notifying Handlers

These Ansible roles and training materials were last tested on Centos 7 and Ubuntu 18.04, but will probably work on other RHEL and Debian variants.

The roles that are used in these training are currently used by usegalaxy.*, and other, servers in maintaining their infrastructure. (US, EU, both are running CentOS 7)

If you have an issue running these trainings on your OS flavour, please report the issue in the training material and we can see if it is possible to solve.

What is Ansible?

Ansible runs commands on local or remote computers. It can move files around, create files from templates, and run command line tools. Primarily used for system administration tasks at scale. It has a push model rather than a pull model like Puppet. If you’ve used Puppet, Ansible doesn’t evaluate what changes need to be made and make those, it just runs through all of commands every time.

Some terms that you should know first:

Inventory file
An Ansible-specific file that defines the systems (“hosts”) and groups of hosts on which Ansible should operate.
Ansible module
A piece of Python code that converts some parameters into an invocation. An example would be the command module which converts parameters like command: ls into a command line that is executed. There are pre-built modules for just about everything.
Task
A call to an Ansible module that should be executed and the configuration for this module.
Role
A folder containing some tasks, templates, files, and default values for variables, with a predefined directory structure. People share roles on “Ansible Galaxy.
Playbook
A YAML file listing a set of tasks and/or roles that should be applied to a group of hosts.
Vault
An encrypted YAML file. You put your secrets here and then you can use them in tasks, roles and playbooks.

Looking at each of these briefly:

Inventory file

[webservers]
192.0.2.0
192.0.2.1

[databases]
192.0.2.3 ansible_user=root

Here we’ve defined two groups of computers, webservers and databases. ansible_user is used to specify which user to connect with (you need to specify that if you SSH in with a username different than your current local user account’s name).

As you can see, in this inventory we connect to multiple remote machines. This is common practice; Ansible playbooks are stored on your laptop or a build server, and from there Ansible connects out to the remote machines that are managed. The advantage of running remotely is that you can manage dozens of machines simultaneously, rather than just the local machine. This scaling out to N machines is one of the strengths of Ansible.

Additionally, playbooks are often stored in Git or other version control repositories to ensure that if anything happens to the infrastructure you manage, you’ll still have a copy of the information required to rebuild everything with Ansible.

For more advanced features of the inventory file, check out the official documentation on this topic.

Roles

We can look at the ansible-cvmfs role as an example for the layout of a role:

├── defaults
│   └── main.yml
├── files
│   ├── cvmfs_remount_sync.c
│   ├── cvmfs_remount_sync.centos_6
│   ├── cvmfs_remount_sync.centos_7
│   └── ...
├── handlers
│   └── main.yml
├── meta
│   └── main.yml
├── tasks
│   ├── apache.yml
│   ├── ...
│   ├── main.yml
│   └── ...
├── templates
│   ├── ...
│   └── stratum1_squid.conf.j2
└── vars
    ├── ...
    ├── main.yml
    └── ...

These are the folders that are included in many complex roles. Simpler roles will often not need all of the folders.

Folder Usage
defaults Default values for variables the user can set (e.g. “version of software to install”).
files These are files which should be copied as-is over to the remote location.
handlers This is typically used for restarting processes.
meta Only needed to publish your role to Ansible Galaxy.
tasks Always start reading here. This is the most important folder and the best place to start when trying to understand what an unfamiliar role does. Anything that is loaded will be referenced here (e.g. variables to load, handlers, files, templates).
templates Files that are templated out with variables before being copied.
vars Default values for variables the user should normally not change (e.g. name of a package in different Linux distributions).

For more information check out the official documentation on this topic.

Modules and Tasks

A tasks/main.yml file calls multiple Ansible modules to accomplish its goal. A typical tasks file looks like:

---
- name: Download cvmfs_preload utility when desired
  get_url:
    url: https://cvmrepo.web.cern.ch/cvmrepo/preload/cvmfs_preload
    dest: "{{ cvmfs_preload_path }}/cvmfs_preload"
    owner: root
    group: root
    mode: 755
  when: cvmfs_preload_install | bool

- name: Ensure AutoFS is enabled + running
  service:
    name: autofs
    enabled: yes
    state: started

Here we have two tasks. Each has a name that will be shown to the person running the playbook.

The first example invokes the get_url module with 5 arguments (url, dest, owner, group and mode). This will download a file from the location indicated in url to the directory specified by dest, change user and group ownerships to root, and change file permissions to 755 (as assigned by the mode argument). The dest parameter uses a Jinja2 template to evaluate the value of the variable cvmfs_preload_path. We can grep through the repository and see that defaults/main.yml sets that to /usr/bin by default. We can override this if we need, we’ll come back to that later. The first task also has a when condition to ensure it only runs when the cvmfs_preload_install variable is set.

The second invokes the service module. The arguments to this one are quite legible and the functionality can be inferred from the names for the most part: The service name: autofs will be enabled and its state should be started.

Many modules are available for you to use.

Stylistic Choices

Ansible accepts two ways to format tasks:

YAML style:

- package:
    name: "{{ package_name }}"
    state: "{{ package_state }}"

And an inline style.

- package name={{ package_name }} state={{ package_state }}

Some groups prefer one style or another. You can mix both of these but you probably shouldn’t. In the YAML style the templated value needs to be quoted if the value after the colon starts with a { (see https://docs.ansible.com/ansible/latest/user_guide/playbooks_variables.html#when-to-quote-variables-a-yaml-gotcha ). The inline style does not require quoting of templated values.

Playbooks

- name: CVMFS
  hosts: webservers
  vars:
    cvmfs_numfiles: 4096
  roles:
    - galaxyproject.cvmfs

This is a quite minimal playbook. It selects a hosts group named webservers, overrides the variable cvmfs_numfiles, and then says the following set of roles will be executed for this group of hosts. Ansible makes it easy to collect tasks that should apply to a group of hosts and run a playbook for all of those hosts. Some good uses of this are things like ensuring a certain set of users are installed on all of your managed machines, or using one of the package autoupdating roles to make sure your machines are up-to-date.

For more information check out the official documentation on this topic.

Philosophies

Different groups use playbooks differently. Here is a non-exhaustive list of ways that the author has seen playbooks used:

Strategy Use Cases
A single playbook that does everything Works well in cloud use cases where you would rather terminate a VM and just restart from nothing. If 100% of the tasks are idempotent (can be re-run without causing problems) then this can also work well for ensuring all machines that are managed are meeting your compliance standards.
Playbooks that execute one-off commands Works well when you have commands that cannot safely be re-run (e.g. you are managing a repository on a remote machine and executing a ‘commit’ action.) Also works if you have multiple people who are responsible for managing a machine but you want to ensure exactly the same commands are run each time a task needs to be done, with no variation.

Variables

There are a bunch of places variables can be set. The list is ridiculous. Try and find a convention within your group and stick to that to help manage the chaos.

There are some special places you can put variables that will be loaded automatically:

  • group_vars/<group_name>.yml: if you run a playbook which has hosts: webservers, and group_vars/webservers.yml exists, it will be loaded.
  • host_vars/<host_name>.yml: if you run a playbook and it affects host (e.g.) api01, and host_vars/api01.yml exists, it will be loaded automatically.

For a real-life example, UseGalaxy.eu generally attempts to put variables in an appropriately named group_vars/ file.

Vaults

ansible-vault is a super useful tool that lets you encrypt secrets for your group using a pre-shared key. This allows you to commit ALL of your Ansible playbooks and variables to source control, without the concern of leaking secrets. E.g. UseGalaxy.eu’s vault file. The encrypted version is not very interesting to look at, but it is mostly to show that we confidently place an encrypted copy of our secrets online, under configuration management. This has made our life a lot easier.

Your First Playbook and First Role

The above introduction was certainly not enough for you to feel confident in Ansible, so we will now build a small role and a playbook to run this role.

Warning: Safety First

Many of the things you can do with Ansible can be quite dangerous. As dangerous as normally being at the command line, but scaled across N machines. Be very careful with the changes you plan to make. Ansible provides some flags which can help you identify changes before they’re made to production systems:

--diff
This will show the difference whenever a file is changed.
--check
Will do a dry-run of the playbook, attempting to show you which tasks will execute, which files will be updated. Not all actions can be predicted because commands are not run, if a downstream task depends on the result of a command execution task, Ansible has no way of knowing whether or not it will execute.

In this tutorial we will write to files in /tmp as that is a relatively safe thing to do. The training material community does not have the resources to test this tutorial across all of the platforms you might want to run it on. Additionally we do not want to be responsible if you accidentally cause permanent damage by following this tutorial.

Comment: Requirements for Running This Tutorial
  1. You have Ansible installed on the machine where you will install Galaxy

    Comment: Running Ansible on remote machine

    It is possible to have Ansible installed on your laptop/local machine and run it against some remote hosts as well. We will not do that in this training.

  2. Your ansible version is >=2.7, you can check this by running ansible --version

A Basic Role

Hands-on: Setting up our workspace
  1. In this training we will run Ansible on the machine it will modify. This is not best practice, but it is convenient for trainings. You should probably run this in a VM either on the Cloud or in VirtualBox or similar

    All of the steps are the same, no matter which machine Ansible will manage and where you run it. The only difference is the connection setup

  2. Create a directory named intro and cd into it.

    It’s good practice to keep your deployments separated, and later on we’ll be deploying Galaxy with ansible in a separate directory.

  3. Create your inventory file (named hosts) in this folder

    1. We will call our group my_hosts

    2. Create an inventory file with the group my_hosts and localhost ansible_connection=local, which tells Ansible to not use SSH, and just use the local connection. Additionally, you should explicitly set the ansible_user variable to the username to use when connecting to the server. Ansible has changed its behaviour over time regarding whether or not ansible_user is defined, and it is most effective to define it explicitly even when it can sometimes be inferred.

      The file should look like:

      [my_hosts]
      localhost ansible_connection=local ansible_user=ubuntu
      

      For more advanced features of the inventory file, check out the official documentation on this topic.

  4. Create the roles directory, your role, and the tasks folder: mkdir -p roles/my-role/tasks/

  5. Create a YAML file in that directory, roles/my-role/tasks/main.yml and open it for editing

  6. Define a copy task like below:

    ---
    - name: Copy a file to the remote host
      copy:
        src: test.txt
        dest: /tmp/test.txt
    

    You can read about all of the parameters available to the copy module in Ansible’s documentation.

    You can usually find a module for most commands you will run in a shell, e.g. by searching the internet for “ansible copy file to server” or “ansible restart service”. If you cannot find a module that does it, there is the command module, but this should be avoided if possible. Expect to have a browser session with 10-30 different Ansible module documentation tabs if you work with Ansible regularly, no one remembers what arguments are available to every module.

  7. Create a roles/my-role/files folder, and within it a file named test.txt, containing the content “Hello, Galaxy 🚀”

  8. This is a complete role by itself and will copy the file test.txt from the roles/my-role/files/ folder over to the target host and place it in /tmp.

  9. Create and open playbook.yml file for editing in the intro directory you created. Place the following content in there:

    ---
    - hosts: my_hosts
      roles:
        - my-role
    
    Question

    How does your file tree look now? Use find or tree.

    .
    ├── hosts
    ├── playbook.yml
    └── roles
        └── my-role
            ├── files
            │   └── test.txt
            └── tasks
                └── main.yml
    
  10. Run the playbook:

    Input: Bash
    ansible-playbook -i hosts playbook.yml
    
    Output: Bash
    PLAY [my_hosts] ****************************************************************
    TASK [Gathering Facts] *********************************************************
    ok: [localhost]
    TASK [my-role : Copy a file to the remote host] ********************************
    changed: [localhost]
    PLAY RECAP *********************************************************************
    localhost                  : ok=2    changed=1    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0
    
  11. Login to the appropriate host (unless your host was localhost) and cat /tmp/test.txt to see that the change was made.

Now that you’ve done this, here are some starting points for exploration:

  • Add more hosts, watch as Ansible executes over all of them in parallel.
  • Identify a task you do regularly, e.g. restarting a service. Find the Ansible service module and add that to your playbook.
Comment: Too Many Cows?

If you’ve installed the cowsay tool, Ansible (for some reason) will take advantage of that to output a lot of the output with cowsay. To disable this you can export ANSIBLE_NOCOWS=1 (Remember that exporting will only last as long as the current invocation of your terminal does, so consider adding this to your user profile if you wish to keep cowsay installed and still have legible output.)

 ________________
< PLAY [my_hosts] >
 ----------------
        \   ^__^
         \  (oo)\_______
            (__)\       )\/\
                ||----w |
                ||     ||
 ________________________
< TASK [Gathering Facts] >
 ------------------------
        \   ^__^
         \  (oo)\_______
            (__)\       )\/\
                ||----w |
                ||     ||

Facts

In the last step of the last hands-on, you ran your playbook. The first TASK that was executed was not one that you had written, it reads:

TASK [Gathering Facts] *********************************************************
ok: [localhost]

The setup module runs by default for every host and gathers facts about the host.

Hands-on: The Setup Module
  1. Run the command ansible -i hosts -m setup my_hosts | less.

    The my_hosts at the end of the command refers to the group we defined in our hosts inventory file.

  2. Investigate the output. See what sort of information is made available to you.

    Question
    1. What variable stores the OS name?
    2. Where are all of the places that you can find your machine’s IP (v4) addresses?
    1. The OS name is stored in ansible_distribution. We saw ansible_os_family used above in the ansible-cvmfs role. You can use these variables if you are writing a generic role but packages or commands are named different on different operating systems.

    2. Ansible stores network information in quite a few places, sometimes one place is more convenient or more correct to use:

      • ansible_all_ipv4_addresses
      • ansible_default_ipv4.address
      • ansible_<interface_name>.ipv4

Templates

Templates give you greater control over the files you are deploying to the target system. If you need to deploy a file to multiple hosts, but configure it differently on each host, you should use templates. For instance, deploying a service that should only listen on the correct IP address for that host would be a good use case for templates. All of the facts you discovered in the previous hands-on are available to you to use in templates, when statements (like the ansible-cvmfs example we saw earlier). Additionally all of the variables you’ve defined are available as well.

Templates end with the .j2 suffix and use Jinja2 syntax. If you are not familiar with it, you should read about it first, before moving on with the tutorial. Ansible fills the templates with variable values and copies the file to its remote destination without the .j2 suffix.

Hands-on: Variables and Templates
  1. Create the directories: roles/my-role/templates, roles/my-role/defaults

  2. Create and edit a file for your role’s variables, roles/my-role/defaults/main.yml

  3. Insert the following content:

    ---
    server_name: Cats!
    

    This will define a variable server_name that can then be used in templates.

  4. Create and edit roles/my-role/templates/test.ini.j2

  5. Insert the following content:

    [example]
    server_name = {{ server_name }}
    listen = {{ ansible_default_ipv4.address }}
    
  6. Edit roles/my-role/tasks/main.yml and append a new task to the end to template this file:

    - name: Template the configuration file
      template:
        src: test.ini.j2
        dest: /tmp/test.ini
    
  7. Run the playbook again.

  8. Check the contents of /tmp/test.ini

    Input: Bash
    cat /tmp/test.ini
    
    Output

    The file should look like:

    [example]
    server_name = Cats!
    listen = 192.168.0.2
    

    Where the last line has the machine’s IP address.

    Now that this has worked successfully, we will setup a group_vars folder to show how a person using my-role would override the server_name variable.

  9. Create the folder group_vars/ (in the root of your directory)

  10. Create and edit group_vars/my_hosts.yml

  11. Insert the following:

    ---
    server_name: Dogs!
    

    When the playbook runs, as part of the setup, it collects any variables that are set. For a playbook affecting a group of hosts named my_hosts, it checks many different places for variables, including “group_vars/my_hosts.yml”. If there are variables there, they’re added to the collection of current variables. It also checks “group_vars/all.yml” (for the built-in host group all). There is a precedence order, but then these variables are available for roles and tasks to consume.

  12. Run the playbook again, but imagine you are worried about this change, and supply the --check --diff flag to see what changes are made before committing to make them.

    Input: Bash
    ansible-playbook -i hosts playbook.yml --check --diff
    

    If you forget to use --diff, it is not easy to see what has changed. Some modules like the copy and template modules have a backup option. If you set this option, then it will keep a backup copy next to the destination file.

    However, most modules do not have such an option, so if you want to know what changes, always use --diff.

    Output
    PLAY [my_hosts] ****************************************************************
    TASK [Gathering Facts] *********************************************************
    ok: [localhost]
    TASK [my-role : Copy a file to the remote host] ********************************
    ok: [localhost]
    TASK [my-role : Template the configuration file] *******************************
    --- before: /tmp/test.ini
    +++ after: /home/hxr/.ansible/tmp/ansible-local-1906887dr2u6j8n/tmptx9pdelg/test.ini.j2
    @@ -1,3 +1,3 @@
     [example]
    -server_name = Cats!
    +server_name = Dogs!
     listen = 192.168.0.25
    changed: [localhost]
    PLAY RECAP **********************************************
    localhost                  : ok=3    changed=1    unreachable=0    failed=0    skipped=0    rescued=0    ignored=0
    

    Here you can see that the server_name value will be changed. Despite Ansible reporting changed=1, no changes have actually been applied to the system.

  13. Run the playbook again, without the --check flag to apply your changes.

Comment: Ansible Variable Templating

In this hands-on we used some templated variables. We defined them in a template, but they are also commonly used in the group variables file. Our templated variable looked like: listen = {{ ansible_default_ipv4.address }}.

It is common to see things like this in Ansible roles:

root_dir = /opt/my-app
config_dir = "{{ root_dir }}/config"

When Ansible runs:

  1. It collects variables defined in group variables and other places
  2. The first task for each machine is the setup module which gathers facts about the host, which are added to the available variables
  3. When multiple roles execute in a playbook:
    1. Their defaults are added to the set of variables (the group variables having precedence over these variables)
    2. They can also dynamically define more variables which may not be set until that role is run
  4. Before use (in templates, commands, etc.), variables are resolved to their final value

Ansible Galaxy

Now that you’ve built a small role, you can imagine that building real roles that manage the full installation of a piece of software are not simple things. Ansible Galaxy is the answer here. Many roles for common administration tasks, and software installation and setup are readily available on Ansible Galaxy.

Warning: This will install Memcached on the target machine.

Hands-on: Installing a module using ansible-galaxy
  1. Install the geerlingguy.memcached role with ansible-galaxy into your roles folder:

    Input: Bash
    ansible-galaxy install \
        -p roles/ geerlingguy.memcached
    
    Output

    This will install the new role into your roles folder, alongside your own role.

    - downloading role 'memcached', owned by geerlingguy
    - downloading role from https://github.com/geerlingguy/ansible-role-memcached/archive/1.1.0.tar.gz
    - extracting geerlingguy.memcached to /home/ubuntu/ansible/intro/roles/geerlingguy.memcached
    - geerlingguy.memcached (1.1.0) was installed successfully
    
  2. Edit your playbook.yml and add the role geerlingguy.memcached at the bottom, after my-role

  3. Run the playbook

    Question

    Did something go wrong?

    Since you have been running the playbook as a non-root user (or at least you should have been!), the step to install a package fails. The solution to this is to set become: true.

    1. Edit your playbook.yml
      • add become: true just below hosts: my_hosts
    2. Run the playbook again

    become causes Ansible to attempt to become a different user (using sudo/su/whatever is appropriate), by default this is root. If you want to become a different user, just set become_user. Beware, the user should be able to privilege escalate without a password prompt. Otherwise when you execute the playbook you should set --ask-become-pass, using the privilege escalation password for that host.

    See the documentation if you need to control this behaviour differently. become can be set either at the task level or the playbook level.

    This is often site-specific policy. If your site doesn’t have a dictated policy, then it’s generally preferable to not allow direct login as root, and login as a separate user.

Choosing a Role

Picking the best role for a task from Ansible Galaxy is not always a trivial task. Sometimes there will only be a single role doing what you need. Other times you’ll have to choose between 20 different roles that all look more or less the same. Here are some tips to guide you in identifying appropriate and well-written roles:

  • The name should match the software you are using (I.e. ignore a role named stackstorm when you are trying to set up rabbitmq. Ansible Galaxy does not have perfect search.)
  • geerlingguy wrote a huge number of roles for many pieces of standard software. If there is a role from this person, this is usually a safe choice. In the same way we recommend galaxyproject and usegalaxy_eu roles.
  • Check the GitHub readme of each role you consider using. Look for:
    • Extensive documentation of all of the variables, their default values, and how they behave.
    • An example playbook using the role
  • A large number of downloads is a good sign that other people found it useful

These are usually good proxies for quality, but do not treat them as strict rules. For an example of a role meeting many of these qualities, ansible-cvmfs is good; the variables are well documented and there are example playbooks that you can (more or less) copy-and-paste and run.

Sometimes a role will accomplish 95% of what you need to do, but not everything. Once you have installed the role with ansible-galaxy install, you can edit it locally to make any changes. In an ideal world you would contribute this back, but this is not always a high priority. Many projects copy roles directly into their repositories, e.g. galaxyproject and usegalaxy.eu

You don’t. There is no standard way for reporting this, but well written roles by trusted authors (e.g. geerlingguy, galaxyproject) do it properly and write all of the variables in the README file of the repository. We try to pick sensible roles for you in this course, but, in real life it may not be that simple.

So, definitely check there first, but if they aren’t there, then you’ll need to read through defaults/ and tasks/ and templates/ to figure out what the role does and how you can control and modify it to accomplish your goals.

Ansible Vault

Now that you have a small role built up, you might start thinking about deploying larger and more complex services and infrastructure. One last common task we want to cover here is the inclusion of secrets. Ansible Vault is really useful to include encrypted secrets in your playbook repository.

Hands-on: Setting up secrets
  1. Run mkdir -p secret_group_vars

  2. Now we’ll create a strong password: openssl rand -base64 24 > vault-password.txt

  3. Run ansible-vault create secret_group_vars/all.yml --vault-password-file=vault-password.txt

    This will open secret_group_vars/all.yml in your text editor.

  4. In this file, enter the following contents and save it:

    apikey: super-secret-api-key-wow!
    
    Question

    How does your file look? Is it readable? Run cat secret_group_vars/all.yml

    The file will look like this, it is encrypted by Ansible Vault with AES256 encryption.

    $ cat secret_group_vars/all.yml
    $ANSIBLE_VAULT;1.1;AES256
    64373665366130333437393639343534653134346538636239393363373062393830653333323966
    3134333366363130326139323162323131643763336236320a393262303938316262643764323862
    36393161666663353231366336613838633866323230303031313465646333613862363264323139
    3263383530626262370a666139666462663938343531656432353239346532316630366165376566
    34313765353766666330366632303836353863396430343264303032363739666139383830323565
    6133663637356331613062353834646561653366386665623930
    
  5. Use the new variable in our .ini file from earlier. Edit roles/my-role/templates/test.ini.j2 and add the line apikey = {{ apikey }}

  6. Add encrypted variable file to playbook.yml

    - hosts: my_hosts
      ...
      vars_files:
       - secret_group_vars/all.yml
    
  7. Tell ansible where to find the decryption file. Create file ansible.cfg with content

    [defaults]
    vault_password_file=vault-password.txt
    
  8. Run the playbook

  9. Check the contents of /tmp/test.ini

    Input: Bash
    cat /tmp/test.ini
    
    Output

    The file should look like:

    [example]
    server_name = Dogs!
    listen = 192.168.0.2
    apikey = super-secret-api-key-wow!
    

In real life scenarios where you are sharing your playbooks publicly, be sure to encrypt all secrets from the start (or fix/remove the git history if you ever did.) If you are storing your vault password in a file, remember to add it to your .gitignore (or VCS appropriate file.)

Other Stuff

Ansible has a huge array of features and we can’t cover them all. Some commonly used features are documented below:

With Items

Duplicating tasks ten times to install ten packages is not efficient, so Ansible provides loop construct

- name: Install stuff
  package:
    name: "{{ item }}"
    state: installed
  loop:
    - htop
    - git
    - vim

This works for any task, not just package installation, if you have things you’d like to repeat. Note that for this task one would normally list multiple packages in the name parameter of the package module.

- name: Install stuff
  package:
    name:
      - htop
      - git
      - vim
    state: installed

When Changed

Doing something only when a task is in the “changed” state is a common pattern. This is often used for reloading a service when some configuration files have changed.

- name: Copy configuration file
  copy:
    src: main.conf
    dest: /etc/service/main.conf
    owner: root
    group: root
    mode: 0644
  register: service_conf

- name: Restart the service
  service:
    name: service
    enabled: yes
    state: restarted
  when: service_conf is changed

Notifying Handlers

Often you want to restart a service whenever something has changed, like above. But if you need to restart the service whenever one of many different tasks has changed, there is a simpler way than writing when: a.changed or b.changed or c.changed or ...

First, move the restarting or reloading of the service into the handlers/main.yml

- name: restart service
  service:
    name: service
    enabled: yes
    state: restarted

- name: reload httpd
  service:
    name: httpd
    enabled: yes
    state: reloaded

Now you can change your task definitions:

- name: Copy configuration file
  copy:
    src: main.conf
    dest: "/etc/service/main.conf
    owner: "root"
    group: "root"
    mode: "0644"
  notify: 'restart service'

We no longer register the command, instead the notify attribute automatically marks that the appropriate handler should be called at the end of the playbook run. Additionally, tasks from outside of this role can use the same notify. So if you use a community built role for managing apache, and then have a custom role that builds the apache configuration, you can use changes there to reload the service using their definition of the reload/restart.

Key points
  • Ansible lets you do system administration at scale

  • Many system administration, software installation, and software management tasks are already available as Ansible tasks or roles

Frequently Asked Questions

Have questions about this tutorial? Check out the tutorial FAQ page or the FAQ page for the Galaxy Server administration topic to see if your question is listed there. If not, please ask your question on the GTN Gitter Channel or the Galaxy Help Forum

Feedback

Did you use this material as an instructor? Feel free to give us feedback on how it went.
Did you use this material as a learner or student? Click the form below to leave feedback.

Click here to load Google feedback frame

Citing this Tutorial

  1. Helena Rasche, Saskia Hiltemann, 2022 Ansible (Galaxy Training Materials). https://training.galaxyproject.org/training-material/topics/admin/tutorials/ansible/tutorial.html Online; accessed TODAY
  2. Batut et al., 2018 Community-Driven Data Analysis Training for Biology Cell Systems 10.1016/j.cels.2018.05.012


@misc{admin-ansible,
author = "Helena Rasche and Saskia Hiltemann",
title = "Ansible (Galaxy Training Materials)",
year = "2022",
month = "10",
day = "18"
url = "\url{https://training.galaxyproject.org/training-material/topics/admin/tutorials/ansible/tutorial.html}",
note = "[Online; accessed TODAY]"
}
@article{Batut_2018,
    doi = {10.1016/j.cels.2018.05.012},
    url = {https://doi.org/10.1016%2Fj.cels.2018.05.012},
    year = 2018,
    month = {jun},
    publisher = {Elsevier {BV}},
    volume = {6},
    number = {6},
    pages = {752--758.e1},
    author = {B{\'{e}}r{\'{e}}nice Batut and Saskia Hiltemann and Andrea Bagnacani and Dannon Baker and Vivek Bhardwaj and Clemens Blank and Anthony Bretaudeau and Loraine Brillet-Gu{\'{e}}guen and Martin {\v{C}}ech and John Chilton and Dave Clements and Olivia Doppelt-Azeroual and Anika Erxleben and Mallory Ann Freeberg and Simon Gladman and Youri Hoogstrate and Hans-Rudolf Hotz and Torsten Houwaart and Pratik Jagtap and Delphine Larivi{\`{e}}re and Gildas Le Corguill{\'{e}} and Thomas Manke and Fabien Mareuil and Fidel Ram{\'{\i}}rez and Devon Ryan and Florian Christoph Sigloch and Nicola Soranzo and Joachim Wolff and Pavankumar Videm and Markus Wolfien and Aisanjiang Wubuli and Dilmurat Yusuf and James Taylor and Rolf Backofen and Anton Nekrutenko and Björn Grüning},
    title = {Community-Driven Data Analysis Training for Biology},
    journal = {Cell Systems}
}
                   

Congratulations on successfully completing this tutorial!