nfc_glpi/docs/projects/nfc_glpi/config_from_code.md

title, description, date, template, about

title	description	date	template	about
Config as code	How to use No Fuss Computings Ansible role to configure GLPI from configuration as code.	2023-07-28	project.html	https://gitlab.com/nofusscomputing/projects/ansible/nfc_glpi

To configure GLPI using this role, the config/settings are saved as json files. The workflow to create the json files, is to (preferably within a test environment) make the required setting changes in GLPI using the GUI. one the changes have been made, conduct an API GET query to GLPI for the item you changed. The JSON body that is returned, is placed within the Config file (detailed below) under path .body. You must however, remove any fields that are not required, for example dynamic fields (fields that update themselves or are auto created) for example date fields. it's also recommended to remove the items id entry from the body, this option enables the updating and creation of the item by the name field.

Config files Example

Each tab provides an example of the JSON file layout, including any additional variables. Any variable listed at path . that are empty strings, you would complete with the items name as it is exactly within GLPI. These fields are used to fetch the item_id of the field by name. This enables having the config saved in a more human readable format. For example, for the field entities_id you would complete it as "entities_id": "My entity name".

!!! tip If you leave the item ID in the config file under path .body.id, the item that matches this ID, will be updated, not the item matching the name.

=== "AuthLDAP"

``` json title="example.json" linenums="1"

{
  "api_path": "AuthLDAP",
  "body": {
    // JSON from API Query
  }
}

```

=== "Config (GLPI General Settings)"

This config file contains the settings found in `General Settings`. The full API JSON response is added to the body. When this file is processed, each individual item is added via the API without using the `id` as found in the JSON file. This is done as the ID's can change, so the workflow includes a step that matches the config item name to the found id in the database, then patches with the values from the JSON file.

``` json title="example.json" linenums="1"
{
  "api_path": "Config",
  "body": [
    // JSON from API Query the full response (see note below)
  ]
}
```

!!! info
    The structure of this JSON file is different than the rest. The body is a **list** of **ALL** of the Config options. This includes the item ID. it was done this way, so that only one API query is required to export all config options.

=== "Entity"

``` json title="example.json" linenums="1"

{
  "api_path": "Entity",
  "entities_id": "",
  "body": {
    // JSON from API Query
  }
}

```

=== "ITILCategory"

``` json title="example.json" linenums="1"

{
  "api_path": "ITILCategory",
  "users_id": "",
  "itilcategories_id": "",
  "entities_id": "",
  "tickettemplates_id_demand": "",
  "body": {
    // JSON from API Query
  }
}

```

=== "Profile"

``` json title="example.json" linenums="1"

{
  "api_path": "Profile",
  "body": {
    // JSON from API Query
  }
}

```

With the config files structured as per the examples, within your playbook(s), load the json files in a list of json objects using variable glpi_config_as_code_json

example

glpi_config_as_code_json:
  - "{{ lookup('file', '/workdir/files/glpi/ITILCategory-new-ldap-user.json') | from_json }}"
  - "{{ lookup('file', '/workdir/files/glpi/LDAPAuth-local_ldap.json') | from_json }}"