Ansible Configuration Precedence

Ansbile loads its configuration from multiple locations/files.

You can get the config file options in configuration from docs.ansible.com and then click on configuration file.

Below it the precendence.

1. env. Variable $ANSIBLE_CONFIG

2. from current directory file: ./ansible.cfg

3. from home directory hidden file: ~/.ansible.cfg

4. global config file: / etc/ ansible/ansible.cfg

The global file comes by default with pip and package manager but not with source compilation, we need to create if we want to use it.

Configuration files are not merges, ansible stops looking when it finds the first result.It won’t merge the values from multiple files.Unlike variable files.

We can override setting in environment variable.

we can export the variables at commandline on the fly using export line or using the profile scripts.

to validate the system to manage is valid and right path.

if we set it false, then we don’t need to add the fingerprints to ssh keys.

Sample config file in current directory, ansible.cfg

[defaults]

host_key_checking=False

And, as there is nothing in ~/.ssh/known_hosts file, the above false value should reflect and execute the ansible commands on remote machines.

ansible web1 -i inventory_prod -m ping

Instead of skipping the fingerprint cheking, it just clicks yes for fingerprint prompts in the ssh prompt in the background.So, it populates the data in the known_hosts file during command execution.

To skip the value from config and get it from env. Variable.In the command prompt.

export $ANSIBLE_HOST_KEY__CHECKING=True

Now, below command will fail.

ansible web1 -i inventory_prod -m ping