This tutorial gets you started with the Inmanta orchestration tool.

Inmanta is intended to manage complex infrastructures, often in the cloud or other virtualized environments. In this guide, we go for a less complex setup: install the Drupal CMS on two VM-like containers. First, we use Docker to set up a basic environment with two empty VM-like containers, an Inmanta server and a postgres server used by inmanta as a database. Then, we use Inmanta to install Drupal on these VM-like containers.


This is meant to get an example Inmanta environment set up and running quickly to experiment with. It is not recommended to run this setup in production, as it might lead to instabilities in the long term.

Setting up the tutorial

To quickly get started with Inmanta, use Docker Compose to set up an environment to host the Inmanta server and some machines to be managed. Before starting this tutorial, first install Docker on your machine. Next install Docker Compose on your machine.

Then, grab the Docker quickstart from our Git repository.

git clone https://github.com/inmanta/quickstart-docker.git
cd quickstart-docker

Now that we have the needed docker files, we will need to get the Inmanta quickstart project itself:

git clone https://github.com/inmanta/quickstart.git quickstart-project

The quickstart project can now be found under the newly created quickstart-project directory. It will be the basis for this quickstart. The quickstart-project directory will also be shared with the Inmanta server container (mounted to /home/inmanta/quickstart-project). We will come back to the files in this repository later.


If you are on Windows, be sure you make the drive with the quickstart project shareable with docker containers:

  1. In Powershell: $env:COMPOSE_CONVERT_WINDOWS_PATHS = 1

  2. Restart Docker for Windows

  3. Go to Docker for Windows settings > Shared Drives > Reset credentials > select drive with quickstart project > set your credentials > Apply

Finally, have Docker Compose deploy the quickstart environment:

docker-compose up

Docker Compose will set up the Inmanta server, a postgres server and two VM-like containers to experiment on. When Docker Compose is done deploying and the Inmanta server is running, you will be able to open the dashboard at When you see the following output, the Inmanta server is ready to be used:

inmanta_quickstart_server | inmanta.protocol.rest    DEBUG   Start REST transport
inmanta_quickstart_server | inmanta                  INFO    Server startup complete


docker-compose will lock the current terminal and use it for output from all 4 containers. You will need to open a new terminal to continue with this quickstart

To get an interactive shell on the Inmanta server (this will be needed later):

docker exec -it "inmanta_quickstart_server" bash


The rest of the quickstart guide assumes commands are executed from the root path of the quickstart-docker Git repository, unless noted otherwise.

Breaking down/Resetting the quickstart-docker environment

To fully clean up or reset the environment, run the following commands:

docker-compose down
docker volume prune -f
docker image rmi inmanta-agent inmanta-server

This will give you a clean environment next time you run docker-compose up.

Automatically deploying Drupal

At this point, you can go through the quickstart guide in one of two ways: via the dashboard or via the command line interface. For the CLI, go to the next section. For the Dashboard, go to Using the dashboard.

Single machine deployment using the CLI

To start a new project, all you need is a directory with a project.yml file, defining the parameters like location to search for modules and where to find the server. In this case we will be using the premade quickstart project we cloned in to ./quickstart-project earlier.

That directory contains a project.yml, which looks like this:

name: quickstart
modulepath: libs
downloadpath: libs
repo: https://github.com/inmanta/
description: A quickstart project that installs a drupal website.
    - apache ~= 0.3.1
    - drupal ~= 0.7.1
    - exec ~= 1.1.0
    - ip ~= 1.0.0
    - logging ~= 0.4.1
    - mysql ~= 0.6.0
    - net ~= 0.5.0
    - php ~= 0.3
    - redhat ~= 0.8.0
    - std ~= 0.26.2
    - web ~= 0.2.2
    - yum ~= 0.5.1

The modulepath setting defines that reusable modules will be stored in libs. The repo setting points to one or more Git projects containing Inmanta modules in Git repositories. The requires setting is used to pin versions of modules, otherwise the latest version is used.

In the next section we will use existing modules to deploy a LAMP stack.

Reusing existing modules

We host modules to set up and manage many systems on our Github. These are available under https://github.com/inmanta/.

When you use an import statement in your model, Inmanta downloads these modules and their dependencies automatically.

The configuration model

In this section we will use the configuration concepts defined in the existing modules to set up Drupal on the host named vm1.

First delete the contents of ./quickstart-project/main.cf, then put in the following:

 1import ip
 2import redhat
 3import redhat::epel
 4import apache
 5import mysql
 6import web
 7import drupal
 9# define the machine we want to deploy Drupal on
10vm1=ip::Host(name="vm1", os=redhat::centos7, ip="", remote_agent=true, remote_user="root")
12# add a mysql and apache http server
14mysql_server=mysql::Server(host=vm1, remove_anon_users=true)
16# deploy drupal in that virtual host
18db=mysql::Database(server=mysql_server, name="drupal_test", user="drupal_test", password="Str0ng-P433w0rd")
19drupal::Application(name=name, container=web_server, database=db, admin_user="admin",
20                    admin_password="test", admin_email="[email protected]",
21                    site_name="localhost")
  • Lines 1-7 import all the required packages.

  • Line 10 defines on which machine we want to deploy Drupal.

    • The name attribute is the hostname of the machine, which is later used to determine what configuration needs to be deployed on which machine.

    • The os attribute defines which operating system this server runs. This is used to select the right tools (yum or dnf or apt).

    • The ip attribute is the IP address of this host. At this moment we define this attribute manually, later in this tutorial we let Inmanta discover this automatically.

  • Line 13 deploys an Apache server on our host.

  • Line 14 deploys a Mysql server on our host and removes its anonymous users.

  • Line 17 defines the name (hostname) of the web application.

  • Line 18 defines a database for our Drupal website.

  • Lines 19-21 define the actual Drupal application.

Deploy the configuration model

To deploy the project, we must first register it with the management server by creating a project and an environment. A project is a collection of related environments. (e.g. development, testing, production, qa,…) An environment is associated with a branch in a git repository. This allows the server to recompile the model when the environment changes.

Connect to the terminal of the server-container:

docker exec -it "inmanta_quickstart_server" bash

Then, create the inmanta project and environment:

cd /home/inmanta/quickstart-project
inmanta-cli project create -n test
inmanta-cli environment create -n quickstart-env -p test -r https://github.com/inmanta/quickstart.git -b master --save


The --save option tells inmanta-cli to store the environment config in the .inmanta file. The compiler uses this file to find the server and to export to the right environment.

Install all module dependencies into the project:

inmanta project install

Finally compile the project and deploy it:

inmanta -vvv  export -d

When the model is sent to the server, it will start deploying the configuration. To track progress, you can go to the dashboard, select the test project and then the quickstart-env environment. When the deployment fails for some reason, consult the troubleshooting page to investigate the root cause of the issue.


The -vvv option sets the output of the compiler to very verbose. The -d option instructs the server to immediately start the deploy.

Accessing your new Drupal server

When the installation is done, you can access your new Drupal server at http://localhost:8080/.

Multi-machine deployment using the CLI

The real power of Inmanta becomes apparent when managing more than one machine. In this section we will move the MySQL server from vm1 to a second machine called vm2.

Update the configuration model

A second machine is easily added to the system by adding the definition of the machine to the configuration model and assigning the MySQL server to the new machine.

Update main.cf to the following:

 1import ip
 2import redhat
 3import redhat::epel
 4import apache
 5import mysql
 6import web
 7import drupal
 9# define the machine we want to deploy Drupal on
10vm1=ip::Host(name="vm1", os=redhat::centos7, ip="", remote_agent=true, remote_user="root")
11vm2=ip::Host(name="vm2", os=redhat::centos7, ip="", remote_agent=true, remote_user="root")
13# add a mysql and apache http server
17# deploy drupal in that virtual host
19db=mysql::Database(server=mysql_server, name="drupal_test", user="drupal_test", password="Str0ng-P433w0rd")
20drupal::Application(name=name, container=web_server, database=db, admin_user="admin",
21                    admin_password="test", admin_email="[email protected]", site_name="localhost")

On line 11 the definition of the new machine is added. On line 15 the MySQL server is assigned to vm2.

Deploy the configuration model

To deploy the configuration model, compile the project and deploy it. In the Inmanta server container terminal:

inmanta -vvv export -d

If you browse to the Drupal site again, the database should be empty once more. When the deployment fails for some reason, consult the troubleshooting page to investigate the root cause of the issue.


When moving the database, a new database is created and the content of the old database is not migrated automatically.

Using the dashboard

Inmanta can deploy from the server using only the dashboard. All changes have to go through the repository in this case.

  1. Clone the quickstart project on github (or to another repository location).

  2. Go to the dashboard.

  3. Create a new project with the name test by clicking Add new project.

  4. Go into the new project and create a new environment by clicking Add new environment:

    • Select the test project.

    • Give the environment a name, e.g. env-quickstart.

    • Specify the repo: for example https://github.com/user/quickstart.

    • Specify the branch: master.

  5. Checkout your clone of the quickstart repository and make changes to the main.cf file, for example add the contents of single_machine.cf to the main.cf file. Commit the changes and push them to your repository.

  6. Go into your new environment.

  7. Press Update & Recompile (this may take a while, as all dependencies are downloaded).

    • Now the Inmanta server downloads the configuration model from your clone of the repository. It also downloads all required modules (i.e. dependencies). These modules contain the instructions to install specific parts of the setup such as for example mysql or drupal itself. To see the source go here, for a more in-depth explanation see above.

    • When this is done, it compiles all modules and integrates them into a new deployment plan.

  8. When the compilation is done, a new version appears. This contains the new deployment plan. Click on this version to open it. This shows a list of all configuration items in this configuration.

  9. Press Deploy to start rolling out this version.

    • An agent is now started that remotely logs in into the virtual machines (via SSH) and starts deploying the Drupal server.

    • It will automatically install the required software and configure it properly.

  10. When the deployment is done, you can find your freshly deployed Drupal instance at http://localhost:8080/.

Create your own modules

Inmanta enables developers of a configuration model to make it modular and reusable. In this section we will create a configuration module that defines how to deploy a LAMP stack with a Drupal site in a two- or three-tiered deployment.

Module layout

A configuration module requires a specific layout:

  • The name of the module is determined by the top-level directory. Within this module directory, a module.yml file has to be specified.

  • The only mandatory subdirectory is the model directory containing a file called _init.cf. What is defined in the _init.cf file is available in the namespace linked with the name of the module. Other files in the model directory create subnamespaces.

  • The files directory contains files that are deployed verbatim to managed machines.

  • The templates directory contains templates that use parameters from the configuration model to generate configuration files.

  • The plugins directory contains Python files that are loaded by the platform and can extend it using the Inmanta API.

|__ module.yml
|__ files
|    |__ file1.txt
|__ model
|    |__ _init.cf
|    |__ services.cf
|__ plugins
|    |__ functions.py
|__ templates
     |__ conf_file.conf.tmpl

We will create our custom module in the libs directory of the quickstart project. Our new module will be called lamp, and we require the _init.cf file (in the model subdirectory) and the module.yml file to have a valid Inmanta module. The following commands create all directories and files to develop a full-featured module:

mkdir ./quickstart-project/libs/{lamp,lamp/model}
touch ./quickstart-project/libs/lamp/model/_init.cf
touch ./quickstart-project/libs/lamp/module.yml


Running into permission errors at this point is normal if you followed the cli version of the quickstart. The best way to resolve these is to sudo mkdir ./quickstart-project/libs/lamp and then sudo chmod -R 777 ./quickstart-project/libs/lamp. Now run the above commands again.

Next, edit the ./quickstart-project/libs/lamp/module.yml file and add meta-data to it:

name: lamp
license: Apache 2.0
version: 0.1

Configuration model

In ./quickstart-project/libs/lamp/model/_init.cf we define the configuration model that defines the lamp configuration module.

 1import ip
 2import apache
 3import mysql
 4import web
 5import drupal
 7entity DrupalStack:
 8    string hostname
 9    string admin_user
10    string admin_password
11    string admin_email
12    string site_name
15index DrupalStack(hostname)
17DrupalStack.webhost [1] -- ip::Host
18DrupalStack.mysqlhost [1] -- ip::Host
20implementation drupalStackImplementation for DrupalStack:
21    # add a mysql and apache http server
22    web_server=apache::Server(host=webhost)
23    mysql_server=mysql::Server(host=mysqlhost)
25    # deploy drupal in that virtual host
26    name=web::Alias(hostname=hostname)
27    db=mysql::Database(server=mysql_server, name="drupal_test", user="drupal_test",
28                       password="Str0ng-P433w0rd")
29    drupal::Application(name=name, container=web_server, database=db, admin_user=admin_user,
30                        admin_password=admin_password, admin_email=admin_email, site_name=site_name)
33implement DrupalStack using drupalStackImplementation
  • Lines 7 to 13 define an entity which is the definition of a concept in the configuration model. On lines 8 to 12, typed attributes are defined which we can later on use in the implementation of an entity instance.

  • Line 15 defines that hostname is an identifying attribute for instances of the DrupalStack entity. This also means that all instances of DrupalStack need to have a unique hostname attribute.

  • Lines 17 and 18 define a relation between a Host and our DrupalStack entity. The first relation reads as follows:

    • Each DrupalStack instance has exactly one ip::Host instance that is available in the webhost attribute.

    • Each ip::Host has zero or one DrupalStack instances that use the host as a webserver. The DrupalStack instance is available in the drupal_stack_webhost attribute.

  • On lines 20 to 31 an implementation is defined that provides a refinement of the DrupalStack entity. It encapsulates the configuration of a LAMP stack behind the interface of the entity by defining DrupalStack in function of other entities, which on their turn do the same. Inside the implementation the attributes and relations of the entity are available as variables.

  • On line 33, the implement statement links the implementation to the entity.

The composition

With our new LAMP module we can reduce the amount of required configuration code in the ./quickstart-project/main.cf file by using more reusable configuration code. Only three lines of site-specific configuration code are required.

 1import ip
 2import redhat
 3import redhat::epel
 4import lamp
 6# define the machine we want to deploy Drupal on
 7vm1=ip::Host(name="vm1", os=redhat::centos7, ip="", remote_agent=true, remote_user="root")
 8vm2=ip::Host(name="vm2", os=redhat::centos7, ip="", remote_agent=true, remote_user="root")
10lamp::DrupalStack(webhost=vm1, mysqlhost=vm2, hostname="localhost", admin_user="admin",
11                  admin_password="test", admin_email="[email protected]", site_name="localhost")

Deploy the changes

Deploy the changes as before, by connection to the servers terminal. Nothing will change because the generated configuration should be exactly the same.

inmanta -vvv export -d

When the deployment fails for some reason, consult the troubleshooting page to investigate the root cause of the issue.