The “Vault” is a feature of Ansible that allows you to keep sensitive data such as passwords or keys protected at rest, rather than as plaintext in playbooks or roles. These vaults can then be distributed or placed in source control.
There are 2 types of vaulted content and each has their own uses and limitations:
Vaulted files: |
|
---|---|
Single encrypted variable: | |
|
Warning
To enable this feature, a command line tool, ansible-vault is used to edit files, and a command line flag --ask-vault-pass
, --vault-password-file
or --vault-id
is used. You can also modify your ansible.cfg
file to specify the location of a password file or configure Ansible to always prompt for the password. These options require no command line flag usage.
For best practices advice, refer to Variables and Vaults.
To run a playbook that contains vault-encrypted data files, you must provide the vault password.
To specify the vault-password interactively:
ansible-playbook site.yml --ask-vault-pass
This prompt will then be used to decrypt (in memory only) any vault encrypted files that are accessed.
Alternatively, passwords can be specified with a file or a script (the script version will require Ansible 1.7 or later). When using this flag, ensure permissions on the file are such that no one else can access your key and do not add your key to source control:
ansible-playbook site.yml --vault-password-file ~/.vault_pass.txt ansible-playbook site.yml --vault-password-file ~/.vault_pass.py
The password should be a string stored as a single line in the file.
If you are using a script instead of a flat file, ensure that it is marked as executable, and that the password is printed to standard output. If your script needs to prompt for data, prompts can be sent to standard error.
Note
You can also set ANSIBLE_VAULT_PASSWORD_FILE
environment variable, e.g. ANSIBLE_VAULT_PASSWORD_FILE=~/.vault_pass.txt
and Ansible will automatically search for the password in that file.
This is something you may wish to do if using Ansible from a continuous integration system like Jenkins.
The --vault-password-file
option can also be used with the ansible-pull command if you wish, though this would require distributing the keys to your nodes, so understand the implications – vault is more intended for push mode.
Ansible 2.4 and later support the concept of multiple vaults that are encrypted with different passwords Different vaults can be given a label to distinguish them (generally values like dev, prod etc.).
The --ask-vault-pass
and --vault-password-file
options can be used as long as only a single password is needed for any given run.
Alternatively the --vault-id
option can be used to provide the password and indicate which vault label it’s for. This can be clearer when multiple vaults are used within a single inventory. For example:
To be prompted for the ‘dev’ password:
ansible-playbook site.yml --vault-id dev@prompt
To get the ‘dev’ password from a file or script:
ansible-playbook site.yml --vault-id dev@~/.vault_pass.txt ansible-playbook site.yml --vault-id dev@~/.vault_pass.py
If multiple vault passwords are required for a single run, --vault-id
must be used as it can be specified multiple times to provide the multiple passwords. For example:
To read the ‘dev’ password from a file and prompt for the ‘prod’ password:
ansible-playbook site.yml --vault-id dev@~/.vault_pass.txt --vault-id prod@prompt
The --ask-vault-pass
or --vault-password-file
options can be used to specify one of the passwords, but it’s generally cleaner to avoid mixing these with --vault-id
.
Note
By default the vault label (dev, prod etc.) is just a hint. Ansible will try to decrypt each vault with every provided password.
Setting the config option DEFAULT_VAULT_ID_MATCH will change this behavior so that each password is only used to decrypt data that was encrypted with the same label. See Labelling Vaults for more details.
Ansible 2.5 and later support using a single executable script to get different passwords depending on the vault label. These client scripts must have a file name that ends with -client
. For example:
To get the dev password from the system keyring using the contrib/vault/vault-keyring-client.py
script:
ansible-playbook --vault-id dev@contrib/vault/vault-keyring-client.py
See Vault Password Client Scripts for a complete explanation of this topic.
As of version 2.3, Ansible can now use a vaulted variable that lives in an otherwise ‘clear text’ YAML file:
notsecret: myvalue mysecret: !vault | $ANSIBLE_VAULT;1.1;AES256 66386439653236336462626566653063336164663966303231363934653561363964363833313662 6431626536303530376336343832656537303632313433360a626438346336353331386135323734 62656361653630373231613662633962316233633936396165386439616533353965373339616234 3430613539666330390a313736323265656432366236633330313963326365653937323833366536 34623731376664623134383463316265643436343438623266623965636363326136 other_plain_text: othervalue
To create a vaulted variable, use the ansible-vault encrypt_string command. See Using encrypt_string for details.
This vaulted variable will be decrypted with the supplied vault secret and used as a normal variable. The ansible-vault
command line supports stdin and stdout for encrypting data on the fly, which can be used from your favorite editor to create these vaulted variables; you just have to be sure to add the !vault
tag so both Ansible and YAML are aware of the need to decrypt. The |
is also required, as vault encryption results in a multi-line string.
Note
Inline vaults ONLY work on variables, you cannot use directly on a task’s options.
This command will output a string in the above format ready to be included in a YAML file. The string to encrypt can be provided via stdin, command line arguments, or via an interactive prompt.
See Use encrypt_string to create encrypted variables to embed in yaml.
© 2012–2018 Michael DeHaan
© 2018–2019 Red Hat, Inc.
Licensed under the GNU General Public License version 3.
https://docs.ansible.com/ansible/2.9/user_guide/playbooks_vault.html