16
4

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?

More than 5 years have passed since last update.

我が家のAnsibleは突然の死にあこがれる

Last updated at Posted at 2017-12-09

はじめに

Ansible Advent Calendar 2017 、10日目の記事です。
ネタ枠です。

Ansibleのタスク出力に cowsay が使えることはよく知られていると思います。

__________________ 
< PLAY [localhost] >
------------------ 
       \   ^__^
        \  (oo)\_______
           (__)\       )\/\
               ||----w |
               ||     ||

_________________________  
< TASK [Gathering Facts] >
------------------------- 
       \   ^__^
        \  (oo)\_______
           (__)\       )\/\
               ||----w |
               ||     ||

cowsayをインストールすると特に何を有効にするでもなくAnsibleのメッセージがcowsay化されてしまい、cowsayを使えると知らない人は パニック に陥ることがあるかもしれません。 ちなみに2年前の私です。

そんなAnsibleのcowsay出力なのですが、cowsayと並ぶおもしろコマンド、 echo-sd で出力をやらせてみることにしました。

ちなみにこの記事でやった方法はコアなソースコードを臆せず触っていく、いわゆる 脳筋プレイ です。
本来であれば以下の記事のように、 Callback Plugins の仕組みを使ってやるのがAnsibleのデザインに沿っています。

どこをいじるか

やりたいことは「 Ansibleがcowsayしているところを突き止め、echo-sdに差し替える 」ことです。
というわけなので、Ansibleコードを覗き見て、cowsayとの繋ぎ位置を突き止めてみます。
OSSなのでソースは 見放題 & 改造し放題 、すばらしいですね。

ansible/ansible - GitHub

とりあえずリポジトリ内検索で素直に「cowsay」と検索してみましょう。

Search - cowsay

ヒットしたファイルは4つ。

ほとんど見るまでもなさそうですね。
十中八九 display.py が犯人でしょう。

display.py
b_COW_PATHS = (
    b"/usr/bin/cowsay",
    b"/usr/games/cowsay",
    b"/usr/local/bin/cowsay",  # BSD path for cowsay
    b"/opt/local/bin/cowsay",  # MacPorts path for cowsay
)

こんなのが含まれているので犯人で間違いなさそうです。
OSごとで微妙にcowsayコマンドのパスが違うのを吸収するようになっているようです。すばらしい。

改造の準備

改造のため、ソースを取得してきます。コードベースには最新安定版の v2.4.2.0-1 を使います。

# Ansibleリポジトリをクローン
$ git clone https://github.com/ansible/ansible.git
$ cd ansible
$ git checkout v2.4.2.0-1

# 実行環境をセットアップ
$ source ./hacking/env-setup
$ ansible --version
ansible 2.4.2.0 (detached HEAD e3a8bf02ac) last updated 2017/12/10 02:06:21 (GMT +900)
  config file = None
  configured module search path = [u'/root/.ansible/plugins/modules', u'/usr/share/ansible/plugins/modules']
  ansible python module location = /opt/ansible/lib/ansible
  executable location = /opt/ansible/bin/ansible
  python version = 2.7.5 (default, Nov  6 2016, 00:28:07) [GCC 4.8.5 20150623 (Red Hat 4.8.5-11)]

実行確認を行うPlaybookはこちら。

echo-sd.yml
- hosts: localhost
  user: ansible
  tasks:
    - name: Display message
      debug:
        msg: "突然の死"

これが通常時のメッセージです。

$ ansible-playbook -i hosts echo-sd.yml

PLAY [localhost] *********************************************************************************************

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

TASK [Display message] ***************************************************************************************
ok: [localhost] => {
    "msg": "突然の死"
}

PLAY RECAP ***************************************************************************************************
localhost                  : ok=2    changed=0    unreachable=0    failed=0

どういじるか

非常に短絡的ですが、こんな感じでcowsayの代わりにecho-sdをぶち込んでみたらどうなるでしょうか。

display.py
b_COW_PATHS = (
    b"/usr/local/bin/echo-sd", # echo-sdのパス
)

上手くいったら儲けもの、祈る気持ちでレッツトライ...!

$ ansible-playbook -i hosts echo-sd.yml


ok: [localhost]

ok: [localhost] => {
    "msg": "echo-sd !!!"
}

localhost                  : ok=2    changed=0    unreachable=0    failed=0

$ echo-sd 何も出ない
_人人人人人人人_
> 何も出ない <
 ̄Y^Y^Y^Y^Y^Y^Y^ ̄

まぁ、上手くいかないこと自体は予想通りです。真面目にやりましょう。

完成版

先に完成版からです。結果的に2つのファイルをいじることになりました。解説は後で順を追ってしていくので、とりあえず折りたたんでおきます。

lib/ansible/utils/display.py
# (c) 2014, Michael DeHaan <michael.dehaan@gmail.com>
#
# This file is part of Ansible
#
# Ansible is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# Ansible is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with Ansible.  If not, see <http://www.gnu.org/licenses/>.

from __future__ import (absolute_import, division, print_function)
__metaclass__ = type

import errno
import fcntl
import getpass
import locale
import logging
import os
import random
import subprocess
import sys
import textwrap
import time

from struct import unpack, pack
from termios import TIOCGWINSZ

from ansible import constants as C
from ansible.errors import AnsibleError
from ansible.module_utils._text import to_bytes, to_text
from ansible.utils.color import stringc


try:
    # Python 2
    input = raw_input
except NameError:
    # Python 3, we already have raw_input
    pass


logger = None
# TODO: make this a logging callback instead
if C.DEFAULT_LOG_PATH:
    path = C.DEFAULT_LOG_PATH
    if (os.path.exists(path) and os.access(path, os.W_OK)) or os.access(os.path.dirname(path), os.W_OK):
        logging.basicConfig(filename=path, level=logging.DEBUG, format='%(asctime)s %(name)s %(message)s')
        mypid = str(os.getpid())
        user = getpass.getuser()
        logger = logging.getLogger("p=%s u=%s | " % (mypid, user))
    else:
        print("[WARNING]: log file at %s is not writeable and we cannot create it, aborting\n" % path, file=sys.stderr)

b_ECHO_SD_PATHS = (
b"/usr/local/bin/echo-sd", # path for echo-sd
)

b_ECHO_SD_STYLES = (
b"vertical",
b"tanzaku",
b"default",
)

class Display:

    def __init__(self, verbosity=0):

        self.columns = None
        self.verbosity = verbosity

        # list of all deprecation messages to prevent duplicate display
        self._deprecations = {}
        self._warns = {}
        self._errors = {}

        self.b_echo_sd = None
        self.nonsd = C.ANSIBLE_SD_SELECTION

        self.set_echo_sd_info()

        self._set_column_width()

    def set_echo_sd_info(self):
        if not C.ANSIBLE_NOSD:
            for b_echo_sd_path in b_ECHO_SD_PATHS:
                if os.path.exists(b_echo_sd_path):
                    self.b_echo_sd = b_echo_sd_path

    def display(self, msg, color=None, stderr=False, screen_only=False, log_only=False):
        """ Display a message to the user

        Note: msg *must* be a unicode string to prevent UnicodeError tracebacks.
        """

        nocolor = msg
        if color:
            msg = stringc(msg, color)

        if not log_only:
            if not msg.endswith(u'\n'):
                msg2 = msg + u'\n'
            else:
                msg2 = msg

            msg2 = to_bytes(msg2, encoding=self._output_encoding(stderr=stderr))
            if sys.version_info >= (3,):
                # Convert back to text string on python3
                # We first convert to a byte string so that we get rid of
                # characters that are invalid in the user's locale
                msg2 = to_text(msg2, self._output_encoding(stderr=stderr), errors='replace')

            if not stderr:
                fileobj = sys.stdout
            else:
                fileobj = sys.stderr

            fileobj.write(msg2)

            try:
                fileobj.flush()
            except IOError as e:
                # Ignore EPIPE in case fileobj has been prematurely closed, eg.
                # when piping to "head -n1"
                if e.errno != errno.EPIPE:
                    raise

        if logger and not screen_only:
            msg2 = nocolor.lstrip(u'\n')

            msg2 = to_bytes(msg2)
            if sys.version_info >= (3,):
                # Convert back to text string on python3
                # We first convert to a byte string so that we get rid of
                # characters that are invalid in the user's locale
                msg2 = to_text(msg2, self._output_encoding(stderr=stderr))

            if color == C.COLOR_ERROR:
                logger.error(msg2)
            else:
                logger.info(msg2)

    def v(self, msg, host=None):
        return self.verbose(msg, host=host, caplevel=0)

    def vv(self, msg, host=None):
        return self.verbose(msg, host=host, caplevel=1)

    def vvv(self, msg, host=None):
        return self.verbose(msg, host=host, caplevel=2)

    def vvvv(self, msg, host=None):
        return self.verbose(msg, host=host, caplevel=3)

    def vvvvv(self, msg, host=None):
        return self.verbose(msg, host=host, caplevel=4)

    def vvvvvv(self, msg, host=None):
        return self.verbose(msg, host=host, caplevel=5)

    def debug(self, msg):
        if C.DEFAULT_DEBUG:
            self.display("%6d %0.5f: %s" % (os.getpid(), time.time(), msg), color=C.COLOR_DEBUG)

    def verbose(self, msg, host=None, caplevel=2):
        if self.verbosity > caplevel:
            if host is None:
                self.display(msg, color=C.COLOR_VERBOSE)
            else:
                self.display("<%s> %s" % (host, msg), color=C.COLOR_VERBOSE, screen_only=True)

    def deprecated(self, msg, version=None, removed=False):
        ''' used to print out a deprecation message.'''

        if not removed and not C.DEPRECATION_WARNINGS:
            return

        if not removed:
            if version:
                new_msg = "[DEPRECATION WARNING]: %s. This feature will be removed in version %s." % (msg, version)
            else:
                new_msg = "[DEPRECATION WARNING]: %s. This feature will be removed in a future release." % (msg)
            new_msg = new_msg + " Deprecation warnings can be disabled by setting deprecation_warnings=False in ansible.cfg.\n\n"
        else:
            raise AnsibleError("[DEPRECATED]: %s.\nPlease update your playbooks." % msg)

        wrapped = textwrap.wrap(new_msg, self.columns, drop_whitespace=False)
        new_msg = "\n".join(wrapped) + "\n"

        if new_msg not in self._deprecations:
            self.display(new_msg.strip(), color=C.COLOR_DEPRECATE, stderr=True)
            self._deprecations[new_msg] = 1

    def warning(self, msg, formatted=False):

        if not formatted:
            new_msg = "\n[WARNING]: %s" % msg
            wrapped = textwrap.wrap(new_msg, self.columns)
            new_msg = "\n".join(wrapped) + "\n"
        else:
            new_msg = "\n[WARNING]: \n%s" % msg

        if new_msg not in self._warns:
            self.display(new_msg, color=C.COLOR_WARN, stderr=True)
            self._warns[new_msg] = 1

    def system_warning(self, msg):
        if C.SYSTEM_WARNINGS:
            self.warning(msg)

    def banner(self, msg, color=None, sd=True):
        '''
        Prints a header-looking line with echo-sd or stars wit hlength depending on terminal width (3 minimum)
        '''
        if self.b_echo_sd and sd:
            try:
                self.banner_echo_sd(msg)
                return
            except OSError:
                self.warning("somebody cleverly deleted echo-sd or something during the PB run.  heh.")

        msg = msg.strip()
        star_len = self.columns - len(msg)
        if star_len <= 3:
            star_len = 3
        stars = u"*" * star_len
        self.display(u"\n%s %s" % (msg, stars), color=color)

    def banner_echo_sd(self, msg, color=None):
        if u": [" in msg:
            msg = msg.replace(u"[", u"")
            if msg.endswith(u"]"):
                msg = msg[:-1]
        runcmd = [self.b_echo_sd]
        if self.nonsd:
            thesd = self.nonsd
            if thesd == 'random':
                thesd = random.choice(list(b_ECHO_SD_STYLES))
            if thesd != 'default':
                runcmd.append(to_bytes("--" + thesd))
                msg = msg.replace(u"[", u"")
                msg = msg.replace(u"]", u"")
        runcmd.append(to_bytes(msg))
        cmd = subprocess.Popen(runcmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
        (out, err) = cmd.communicate()
        print()
        self.display(u"%s\n" % to_text(out), color=color)

    def error(self, msg, wrap_text=True):
        if wrap_text:
            new_msg = u"\n[ERROR]: %s" % msg
            wrapped = textwrap.wrap(new_msg, self.columns)
            new_msg = u"\n".join(wrapped) + u"\n"
        else:
            new_msg = u"ERROR! %s" % msg
        if new_msg not in self._errors:
            self.display(new_msg, color=C.COLOR_ERROR, stderr=True)
            self._errors[new_msg] = 1

    @staticmethod
    def prompt(msg, private=False):
        prompt_string = to_bytes(msg, encoding=Display._output_encoding())
        if sys.version_info >= (3,):
            # Convert back into text on python3.  We do this double conversion
            # to get rid of characters that are illegal in the user's locale
            prompt_string = to_text(prompt_string)

        if private:
            return getpass.getpass(prompt_string)
        else:
            return input(prompt_string)

    def do_var_prompt(self, varname, private=True, prompt=None, encrypt=None, confirm=False, salt_size=None, salt=None, default=None):

        result = None
        if sys.__stdin__.isatty():

            do_prompt = self.prompt

            if prompt and default is not None:
                msg = "%s [%s]: " % (prompt, default)
            elif prompt:
                msg = "%s: " % prompt
            else:
                msg = 'input for %s: ' % varname

            if confirm:
                while True:
                    result = do_prompt(msg, private)
                    second = do_prompt("confirm " + msg, private)
                    if result == second:
                        break
                    self.display("***** VALUES ENTERED DO NOT MATCH ****")
            else:
                result = do_prompt(msg, private)
        else:
            result = None
            self.warning("Not prompting as we are not in interactive mode")

        # if result is false and default is not None
        if not result and default is not None:
            result = default

        if encrypt:
            # Circular import because encrypt needs a display class
            from ansible.utils.encrypt import do_encrypt
            result = do_encrypt(result, encrypt, salt_size, salt)

        # handle utf-8 chars
        result = to_text(result, errors='surrogate_or_strict')
        return result

    @staticmethod
    def _output_encoding(stderr=False):
        encoding = locale.getpreferredencoding()
        # https://bugs.python.org/issue6202
        # Python2 hardcodes an obsolete value on Mac.  Use MacOSX defaults
        # instead.
        if encoding in ('mac-roman',):
            encoding = 'utf-8'
        return encoding

    def _set_column_width(self):
        if os.isatty(0):
            tty_size = unpack('HHHH', fcntl.ioctl(0, TIOCGWINSZ, pack('HHHH', 0, 0, 0, 0)))[1]
        else:
            tty_size = 0
        self.columns = max(79, tty_size - 1)
lib/ansible/config/base.yml
# Copyright (c) 2017 Ansible Project
# GNU General Public License v3.0+ (see COPYING or https://www.gnu.org/licenses/gpl-3.0.txt)
---
ACCELERATE_CONNECT_TIMEOUT:
  default: 1.0
  description:
    - "This setting controls the timeout for the socket connect call, and should be kept relatively low.
      The connection to the accelerate_port will be attempted 3 times before Ansible will fall back to ssh or paramiko
      (depending on your default connection setting) to try and start the accelerate daemon remotely."
    - "Note, this value can be set to less than one second, however it is probably not a good idea to do so
      unless you are on a very fast and reliable LAN. If you are connecting to systems over the internet, it may be necessary to increase this timeout."
  env: [{name: ACCELERATE_CONNECT_TIMEOUT }]
  ini:
  - {key: accelerate_connect_timeout, section: accelerate}
  type: float
  deprecated:
    why: Removing accelerate as a connection method, settings not needed either.
    version: "2.5"
    alternatives: ssh and paramiko
  version_added: "1.4"
ACCELERATE_DAEMON_TIMEOUT:
  default: 30
  description:
    - This setting controls the timeout for the accelerated daemon, as measured in minutes. The default daemon timeout is 30 minutes.
    - "Prior to 1.6, the timeout was hard-coded from the time of the daemon's launch."
    - For version 1.6+, the timeout is now based on the last activity to the daemon and is configurable via this option.
  env: [{name: ACCELERATE_DAEMON_TIMEOUT}]
  ini:
  - {key: accelerate_daemon_timeout, section: accelerate}
  type: integer
  deprecated:
    why: Removing accelerate as a connection method, settings not needed either.
    version: "2.5"
    alternatives: ssh and paramiko
  version_added: "1.6"
ACCELERATE_KEYS_DIR:
  default: ~/.fireball.keys
  description: ''
  deprecated:
    why: Removing accelerate as a connection method, settings not needed either.
    version: "2.5"
    alternatives: ssh and paramiko
  env: [{name: ACCELERATE_KEYS_DIR}]
  ini:
  - {key: accelerate_keys_dir, section: accelerate}
ACCELERATE_KEYS_DIR_PERMS:
  default: '700'
  description: 'TODO: write it'
  env: [{name: ACCELERATE_KEYS_DIR_PERMS}]
  ini:
  - {key: accelerate_keys_dir_perms, section: accelerate}
  deprecated:
    why: Removing accelerate as a connection method, settings not needed either.
    version: "2.5"
    alternatives: ssh and paramiko
ACCELERATE_KEYS_FILE_PERMS:
  default: '600'
  description: 'TODO: write it'
  env: [{name: ACCELERATE_KEYS_FILE_PERMS}]
  ini:
  - {key: accelerate_keys_file_perms, section: accelerate}
  deprecated:
    why: Removing accelerate as a connection method, settings not needed either.
    version: "2.5"
    alternatives: ssh and paramiko
ACCELERATE_MULTI_KEY:
  default: False
  description: 'TODO: write it'
  env: [{name: ACCELERATE_MULTI_KEY}]
  ini:
  - {key: accelerate_multi_key, section: accelerate}
  type: boolean
  deprecated:
    why: Removing accelerate as a connection method, settings not needed either.
    version: "2.5"
    alternatives: ssh and paramiko
ACCELERATE_PORT:
  default: 5099
  description: 'TODO: write it'
  env: [{name: ACCELERATE_PORT}]
  ini:
  - {key: accelerate_port, section: accelerate}
  type: integer
  deprecated:
    why: Removing accelerate as a connection method, settings not needed either.
    version: "2.5"
    alternatives: ssh and paramiko
ACCELERATE_TIMEOUT:
  default: 30
  description: 'TODO: write it'
  env: [{name: ACCELERATE_TIMEOUT}]
  ini:
  - {key: accelerate_timeout, section: accelerate}
  type: integer
  deprecated:
    why: Removing accelerate as a connection method, settings not needed either.
    version: "2.5"
    alternatives: ssh and paramiko
ALLOW_WORLD_READABLE_TMPFILES:
  name: Allow world readable temporary files
  default: False
  description:
    - This makes the temporary files created on the machine to be world readable and will issue a warning instead of failing the task.
    - It is useful when becoming an unprivileged user.
  env: []
  ini:
  - {key: allow_world_readable_tmpfiles, section: defaults}
  type: boolean
  yaml: {key: defaults.allow_world_readable_tmpfiles}
  version_added: "2.1"
ANSIBLE_SD_SELECTION:
  name: echo-sd style selection
  default: default
  description: This allows you to chose a specific echo-sd style for the banners or use 'random' to cycle through them.
  env: [{name: ANSIBLE_SD_SELECTION}]
  ini:
  - {key: sd_selection, section: defaults}
ANSIBLE_FORCE_COLOR:
  name: Force color output
  default: False
  description: This options forces color mode even when running without a TTY or the "nocolor" setting is True.
  env: [{name: ANSIBLE_FORCE_COLOR}]
  ini:
  - {key: force_color, section: defaults}
  type: boolean
  yaml: {key: display.force_color}
ANSIBLE_NOCOLOR:
  name: Suppress color output
  default: False
  description: This setting allows suppressing colorizing output, which is used to give a better indication of failure and status information.
  env: [{name: ANSIBLE_NOCOLOR}]
  ini:
  - {key: nocolor, section: defaults}
  type: boolean
  yaml: {key: display.nocolor}
ANSIBLE_NOSD:
  name: Suppress echo-sd output
  default: False
  description: If you have echo-sd installed but want to avoid the 'echo-sd' (why????), use this.
  env: [{name: ANSIBLE_NOSD}]
  ini:
  - {key: nosd, section: defaults}
  type: boolean
  yaml: {key: display.i_am_no_fun}
ANSIBLE_PIPELINING:
  name: Connection pipelining
  default: False
  description:
    - Pipelining, if supported by the connection plugin, reduces the number of network operations required to execute a module on the remote server,
      by executing many Ansible modules without actual file transfer.
    - This can result in a very significant performance improvement when enabled.
    - "However this conflicts with privilege escalation (become). For example, when using 'sudo:' operations you must first
      disable 'requiretty' in /etc/sudoers on all managed hosts, which is why it is disabled by default."
  env:
    - name: ANSIBLE_PIPELINING
    - name: ANSIBLE_SSH_PIPELINING
  ini:
  - section: connection
    key: pipelining
  - section: ssh_connection
    key: pipelining
  type: boolean
  yaml: {key: plugins.connection.pipelining}
ANSIBLE_SSH_ARGS:
  # TODO: move to ssh plugin
  default: -C -o ControlMaster=auto -o ControlPersist=60s
  description:
    - If set, this will override the Ansible default ssh arguments.
    - In particular, users may wish to raise the ControlPersist time to encourage performance.  A value of 30 minutes may be appropriate.
    - Be aware that if `-o ControlPath` is set in ssh_args, the control path setting is not used.
  env: [{name: ANSIBLE_SSH_ARGS}]
  ini:
  - {key: ssh_args, section: ssh_connection}
  yaml: {key: ssh_connection.ssh_args}
ANSIBLE_SSH_CONTROL_PATH:
  # TODO: move to ssh plugin
  default: null
  description:
    - This is the location to save ssh's ControlPath sockets, it uses ssh's variable substitution.
    - Since 2.3, if null, ansible will generate a unique hash. Use `%(directory)s` to indicate where to use the control dir path setting.
    - Before 2.3 it defaulted to `control_path=%(directory)s/ansible-ssh-%%h-%%p-%%r`.
    - Be aware that this setting is ignored if `-o ControlPath` is set in ssh args.
  env: [{name: ANSIBLE_SSH_CONTROL_PATH}]
  ini:
  - {key: control_path, section: ssh_connection}
  yaml: {key: ssh_connection.control_path}
ANSIBLE_SSH_CONTROL_PATH_DIR:
  # TODO: move to ssh plugin
  default: ~/.ansible/cp
  description:
    - This sets the directory to use for ssh control path if the control path setting is null.
    - Also, provides the `%(directory)s` variable for the control path setting.
  env: [{name: ANSIBLE_SSH_CONTROL_PATH_DIR}]
  ini:
  - {key: control_path_dir, section: ssh_connection}
  yaml: {key: ssh_connection.control_path_dir}
ANSIBLE_SSH_EXECUTABLE:
  # TODO: move to ssh plugin
  default: ssh
  description:
    - This defines the location of the ssh binary. It defaults to `ssh` which will use the first ssh binary available in $PATH.
    - This option is usually not required, it might be useful when access to system ssh is restricted,
      or when using ssh wrappers to connect to remote hosts.
  env: [{name: ANSIBLE_SSH_EXECUTABLE}]
  ini:
  - {key: ssh_executable, section: ssh_connection}
  yaml: {key: ssh_connection.ssh_executable}
  version_added: "2.2"
ANSIBLE_SSH_RETRIES:
  # TODO: move to ssh plugin
  default: 0
  description: Number of attempts to establish a connection before we give up and report the host as 'UNREACHABLE'
  env: [{name: ANSIBLE_SSH_RETRIES}]
  ini:
  - {key: retries, section: ssh_connection}
  type: integer
  yaml: {key: ssh_connection.retries}
ANY_ERRORS_FATAL:
  name: Make Task failures fatal
  default: False
  description: Sets the default value for the any_errors_fatal keyword, if True, Task failures will be considered fatal errors.
  env:
    - name: ANSIBLE_ANY_ERRORS_FATAL
  ini:
    - section:  defaults
      key: any_errors_fatal
  type: boolean
  yaml: {key: errors.any_task_errors_fatal}
  version_added: "2.4"
BECOME_ALLOW_SAME_USER:
  name: Allow becomming the same user
  default: False
  description: This setting controls if become is skipped when remote user and become user are the same. I.E root sudo to root.
  env: [{name: ANSIBLE_BECOME_ALLOW_SAME_USER}]
  ini:
  - {key: become_allow_same_user, section: privilege_escalation}
  type: boolean
  yaml: {key: privilege_escalation.become_allow_same_user}
CACHE_PLUGIN:
  name: Persistent Cache plugin
  default: memory
  description: Chooses which cache plugin to use, the default 'memory' is ephimeral.
  env: [{name: ANSIBLE_CACHE_PLUGIN}]
  ini:
  - {key: fact_caching, section: defaults}
  yaml: {key: facts.cache.plugin}
CACHE_PLUGIN_CONNECTION:
  name: Cache Plugin URI
  default: ~
  description: Defines connection or path information for the cache plugin
  env: [{name: ANSIBLE_CACHE_PLUGIN_CONNECTION}]
  ini:
  - {key: fact_caching_connection, section: defaults}
  yaml: {key: facts.cache.uri}
CACHE_PLUGIN_PREFIX:
  name: Cache Plugin table prefix
  default: ansible_facts
  description: Prefix to use for cache plugin files/tables
  env: [{name: ANSIBLE_CACHE_PLUGIN_PREFIX}]
  ini:
  - {key: fact_caching_prefix, section: defaults}
  yaml: {key: facts.cache.prefix}
CACHE_PLUGIN_TIMEOUT:
  name: Cache Plugin expiration timeout
  default: 86400
  description: Expiration timeout for the cache plugin data
  env: [{name: ANSIBLE_CACHE_PLUGIN_TIMEOUT}]
  ini:
  - {key: fact_caching_timeout, section: defaults}
  type: integer
  yaml: {key: facts.cache.timeout}
COLOR_CHANGED:
  name: Color for 'changed' task status
  default: yellow
  description: Defines the color to use on 'Changed' task status
  env: [{name: ANSIBLE_COLOR_CHANGED}]
  ini:
  - {key: changed, section: colors}
  yaml: {key: display.colors.changed}
COLOR_DEBUG:
  name: Color for debug statements
  default: dark gray
  description: Defines the color to use when emitting debug messages
  env: [{name: ANSIBLE_COLOR_DEBUG}]
  ini:
  - {key: debug, section: colors}
  yaml: {key: display.colors.debug}
COLOR_DEPRECATE:
  name: Color for deprecation messages
  default: purple
  description: Defines the color to use when emitting deprecation messages
  env: [{name: ANSIBLE_COLOR_DEPRECATE}]
  ini:
  - {key: deprecate, section: colors}
  yaml: {key: display.colors.deprecate}
COLOR_DIFF_ADD:
  name: Color for diff added display
  default: green
  description: Defines the color to use when showing added lines in diffs
  env: [{name: ANSIBLE_COLOR_DIFF_ADD}]
  ini:
  - {key: diff_add, section: colors}
  yaml: {key: display.colors.diff.add}
COLOR_DIFF_LINES:
  name: Color for diff lines display
  default: cyan
  description: Defines the color to use when showing diffs
  env: [{name: ANSIBLE_COLOR_DIFF_LINES}]
  ini:
  - {key: diff_lines, section: colors}
COLOR_DIFF_REMOVE:
  name: Color for diff removed display
  default: red
  description: Defines the color to use when showing removed lines in diffs
  env: [{name: ANSIBLE_COLOR_DIFF_REMOVE}]
  ini:
  - {key: diff_remove, section: colors}
COLOR_ERROR:
  name: Color for error messages
  default: red
  description: Defines the color to use when emitting error messages
  env: [{name: ANSIBLE_COLOR_ERROR}]
  ini:
  - {key: error, section: colors}
  yaml: {key: colors.error}
COLOR_HIGHLIGHT:
  name: Color for highlighting
  default: white
  description: Color used for highlights
  env: [{name: ANSIBLE_COLOR_HIGHLIGHT}]
  ini:
  - {key: highlight, section: colors}
COLOR_OK:
  name: Color for 'ok' task status
  default: green
  description: Defines the color to use when showing 'OK' task status
  env: [{name: ANSIBLE_COLOR_OK}]
  ini:
  - {key: ok, section: colors}
COLOR_SKIP:
  name: Color for 'skip' task status
  default: cyan
  description: Defines the color to use when showing 'Skipped' task status
  env: [{name: ANSIBLE_COLOR_SKIP}]
  ini:
  - {key: skip, section: colors}
COLOR_UNREACHABLE:
  name: Color for 'unreachable' host state
  default: bright red
  description: Defines the color to use on 'Unreachable' status
  env: [{name: ANSIBLE_COLOR_UNREACHABLE}]
  ini:
  - {key: unreachable, section: colors}
COLOR_VERBOSE:
  name: Color for verbose messages
  default: blue
  description: Defines the color to use when emitting verbose messages. i.e those that show with '-v's.
  env: [{name: ANSIBLE_COLOR_VERBOSE}]
  ini:
  - {key: verbose, section: colors}
COLOR_WARN:
  name: Color for warning messages
  default: bright purple
  description: Defines the color to use when emitting warning messages
  env: [{name: ANSIBLE_COLOR_WARN}]
  ini:
  - {key: warn, section: colors}
COMMAND_WARNINGS:
  name: Command module warnings
  default: True
  description:
    - By default Ansible will issue a warning when the shell or command module is used and the command appears to be similar to an existing Ansible module.
    - These warnings can be silenced by adjusting this setting to False. You can also control this at the task level with the module optoin ``warn``.
  env: [{name: ANSIBLE_COMMAND_WARNINGS}]
  ini:
  - {key: command_warnings, section: defaults}
  type: boolean
  version_added: "1.8"
DEFAULT_ACTION_PLUGIN_PATH:
  name: Action plugins path
  default: ~/.ansible/plugins/action:/usr/share/ansible/plugins/action
  description: Colon separated paths in which Ansible will search for Action Plugins.
  env: [{name: ANSIBLE_ACTION_PLUGINS}]
  ini:
  - {key: action_plugins, section: defaults}
  type: pathspec
  yaml: {key: plugins.action.path}
DEFAULT_ALLOW_UNSAFE_LOOKUPS:
  name: Allow unsafe lookups
  default: False
  description:
    - "When enabled, this option allows lookup plugins (whether used in variables as ``{{lookup('foo')}}`` or as a loop as with_foo)
      to return data that is not marked 'unsafe'."
    - By default, such data is marked as unsafe to prevent the templating engine from evaluating any jinja2 templating language,
      as this could represent a security risk.  This option is provided to allow for backwards-compatibility,
      however users should first consider adding allow_unsafe=True to any lookups which may be expected to contain data which may be run
      through the templating engine late
  env: []
  ini:
  - {key: allow_unsafe_lookups, section: defaults}
  type: boolean
  version_added: "2.2.3"
DEFAULT_ASK_PASS:
  name: Ask for the login password
  default: False
  description:
    - This controls whether an Ansible playbook should prompt for a login password.
      If using SSH keys for authentication, you probably do not needed to change this setting.
  env: [{name: ANSIBLE_ASK_PASS}]
  ini:
  - {key: ask_pass, section: defaults}
  type: boolean
  yaml: {key: defaults.ask_pass}
DEFAULT_ASK_SUDO_PASS:
  name: Ask for the sudo password
  default: False
  deprecated:
    why: In favor of become which is a generic framework
    version: "2.8"
    alternatives: become
  description:
    - This controls whether an Ansible playbook should prompt for a sudo password.
  env: [{name: ANSIBLE_ASK_SUDO_PASS}]
  ini:
  - {key: ask_sudo_pass, section: defaults}
  type: boolean
DEFAULT_ASK_SU_PASS:
  name: Ask for the su password
  default: False
  deprecated:
    why: In favor of become which is a generic framework
    version: "2.8"
    alternatives: become
  description:
    - This controls whether an Ansible playbook should prompt for a su password.
  env: [{name: ANSIBLE_ASK_SU_PASS}]
  ini:
  - {key: ask_su_pass, section: defaults}
  type: boolean
DEFAULT_ASK_VAULT_PASS:
  name: Ask for the vault password(s)
  default: False
  description:
    - This controls whether an Ansible playbook should prompt for a vault password.
  env: [{name: ANSIBLE_ASK_VAULT_PASS}]
  ini:
  - {key: ask_vault_pass, section: defaults}
  type: boolean
DEFAULT_BECOME:
  name: Enable privilege escalation (become)
  default: False
  description: Toggles the use of privilege escalation, allowing you to 'become' another user after login.
  env: [{name: ANSIBLE_BECOME}]
  ini:
  - {key: become, section: privilege_escalation}
  type: boolean
DEFAULT_BECOME_ASK_PASS:
  name: Ask for the privelege escalation (become) password
  default: False
  description: Toggle to prompt for privilege escalation password.
  env: [{name: ANSIBLE_BECOME_ASK_PASS}]
  ini:
  - {key: become_ask_pass, section: privilege_escalation}
  type: boolean
DEFAULT_BECOME_METHOD:
  name: Choose privilege escalation method
  default: 'sudo'
  description: Privilege escalation method to use when `become` is enabled.
  env: [{name: ANSIBLE_BECOME_METHOD}]
  ini:
    - {section: privilege_escalation, key: become_method}
DEFAULT_BECOME_EXE:
  name: Choose 'become' executable
  default: ~
  description: 'executable to use for privilege escalation, otherwise Ansible will depend on PATH'
  env: [{name: ANSIBLE_BECOME_EXE}]
  ini:
  - {key: become_exe, section: privilege_escalation}
DEFAULT_BECOME_FLAGS:
  name: Set 'become' executable options
  default: ''
  description: Flags to pass to the privilege escalation executable.
  env: [{name: ANSIBLE_BECOME_FLAGS}]
  ini:
  - {key: become_flags, section: privilege_escalation}
DEFAULT_BECOME_USER:
  # FIXME: should really be blank and make -u passing optional depending on it
  name: Set the user you 'become' via privlege escalation
  default: root
  description: The user your login/remote user 'becomes' when using privilege escalation, most systems will use 'root' when no user is specified.
  env: [{name: ANSIBLE_BECOME_USER}]
  ini:
  - {key: become_user, section: privilege_escalation}
  yaml: {key: become.user}
DEFAULT_CACHE_PLUGIN_PATH:
  name: Cache Plugins Path
  default: ~/.ansible/plugins/cache:/usr/share/ansible/plugins/cache
  description: Colon separated paths in which Ansible will search for Cache Plugins.
  env: [{name: ANSIBLE_CACHE_PLUGINS}]
  ini:
  - {key: cache_plugins, section: defaults}
  type: pathspec
DEFAULT_CALLABLE_WHITELIST:
  name: Template 'callable' whitelist
  default: []
  description: Whitelist of callable methods to be made available to template evaluation
  env: [{name: ANSIBLE_CALLABLE_WHITELIST}]
  ini:
  - {key: callable_whitelist, section: defaults}
  type: list
DEFAULT_CALLBACK_PLUGIN_PATH:
  name: Callback Plugins Path
  default: ~/.ansible/plugins/callback:/usr/share/ansible/plugins/callback
  description: Colon separated paths in which Ansible will search for Callback Plugins.
  env: [{name: ANSIBLE_CALLBACK_PLUGINS}]
  ini:
  - {key: callback_plugins, section: defaults}
  type: pathspec
  yaml: {key: plugins.callback.path}
DEFAULT_CALLBACK_WHITELIST:
  name: Callback Whitelist
  default: []
  description:
    - "List of whitelisted callbacks, not all callbacks need whitelisting,
      but many of those shipped with Ansible do as we don't want them activated by default."
  env: [{name: ANSIBLE_CALLBACK_WHITELIST}]
  ini:
  - {key: callback_whitelist, section: defaults}
  type: list
  yaml: {key: plugins.callback.whitelist}
DEFAULT_CONNECTION_PLUGIN_PATH:
  name: Connection Plugins Path
  default: ~/.ansible/plugins/connection:/usr/share/ansible/plugins/connection
  description: Colon separated paths in which Ansible will search for Connection Plugins.
  env: [{name: ANSIBLE_CONNECTION_PLUGINS}]
  ini:
  - {key: connection_plugins, section: defaults}
  type: pathspec
  yaml: {key: plugins.connection.path}
DEFAULT_DEBUG:
  name: Debug mode
  default: False
  description: Toggles debug output in Ansible, VERY verbose and can hinder multiprocessing.
  env: [{name: ANSIBLE_DEBUG}]
  ini:
  - {key: debug, section: defaults}
  type: boolean
DEFAULT_EXECUTABLE:
  name: Target shell executable
  default: /bin/sh
  description:
    - "This indicates the command to use to spawn a shell under for Ansible's execution needs on a target.
      Users may need to change this in rare instances when shell usage is constrained, but in most cases it may be left as is."
  env: [{name: ANSIBLE_EXECUTABLE}]
  ini:
  - {key: executable, section: defaults}
DEFAULT_FACT_PATH:
  name: local fact path
  default: ~
  description:
    - "This option allows you to globally configure a custom path for 'local_facts' for the implied M(setup) task when using fact gathering."
    - "If not set, it will fallback to the default from the M(setup) module: ``/etc/ansible/facts.d``."
    - "This does **not** affect  user defined tasks that use the M(setup) module."
  env: [{name: ANSIBLE_FACT_PATH}]
  ini:
  - {key: fact_path, section: defaults}
  type: path
  yaml: {key: facts.gathering.fact_path}
DEFAULT_FILTER_PLUGIN_PATH:
  name: Jinja2 Filter Plugins Path
  default: ~/.ansible/plugins/filter:/usr/share/ansible/plugins/filter
  description: Colon separated paths in which Ansible will search for Jinja2 Filter Plugins.
  env: [{name: ANSIBLE_FILTER_PLUGINS}]
  ini:
  - {key: filter_plugins, section: defaults}
  type: pathspec
DEFAULT_FORCE_HANDLERS:
  name: Force handlers to run after failure
  default: False
  description:
    - This option controls if notified handlers run on a host even if a failure occurs on that host.
    - When false, the handlers will not run if a failure has occurred on a host.
    - This can also be set per play or on the command line. See Handlers and Failure for more details.
  env: [{name: ANSIBLE_FORCE_HANDLERS}]
  ini:
  - {key: force_handlers, section: defaults}
  type: boolean
  version_added: "1.9.1"
DEFAULT_FORKS:
  name: Number of task forks
  default: 5
  description: Maximum number of forks Ansible will use to execute tasks on target hosts.
  env: [{name: ANSIBLE_FORKS}]
  ini:
  - {key: forks, section: defaults}
  type: integer
DEFAULT_GATHERING:
  name: Gathering behaviour
  default: 'implicit'
  description:
    - This setting controls the default policy of fact gathering (facts discovered about remote systems).
    - "When 'implicit' (the default), the cache plugin will be ignored and facts will be gathered per play unless 'gather_facts: False' is set."
    - "When 'explicit' the inverse is true, facts will not be gathered unless directly requested in the play."
    - "The 'smart' value means each new host that has no facts discovered will be scanned,
      but if the same host is addressed in multiple plays it will not be contacted again in the playbook run."
    - "This option can be useful for those wishing to save fact gathering time. Both 'smart' and 'explicit' will use the cache plugin."
  env: [{name: ANSIBLE_GATHERING}]
  ini:
    - key: gathering
      section: defaults
  version_added: "1.6"
  choices: ['smart', 'explicit', 'implicit']
DEFAULT_GATHER_SUBSET:
  name: Gather facts subset
  default: 'all'
  description:
      - Set the `gather_subset` option for the M(setup) task in the implicit fact gathering.
        See the module documentation for specifics.
      - "It does **not** apply to user defined M(setup) tasks."
  env: [{name: ANSIBLE_GATHER_SUBSET}]
  ini:
    - key: gather_subset
      section: defaults
  version_added: "2.1"
DEFAULT_GATHER_TIMEOUT:
  name: Gather facts timeout
  default: 10
  description:
    - Set the timeout in seconds for the implicit fact gathering.
    - "It does **not** apply to user defined M(setup) tasks."
  env: [{name: ANSIBLE_GATHER_TIMEOUT}]
  ini:
  - {key: gather_timeout, section: defaults}
  type: integer
  yaml: {key: defaults.gather_timeout}
DEFAULT_HANDLER_INCLUDES_STATIC:
  name: Make handler M(include) static
  default: False
  description:
    - "Since 2.0 M(include) can be 'dynamic', this setting (if True) forces that if the include appears in a ``handlers`` section to be 'static'."
  env: [{name: ANSIBLE_HANDLER_INCLUDES_STATIC}]
  ini:
  - {key: handler_includes_static, section: defaults}
  type: boolean
  deprecated:
    why: include itself is deprecated and this setting will not matter in the future
    version: "2.8"
    alternatives: none as its already built into the decision between include_tasks and import_tasks
DEFAULT_HASH_BEHAVIOUR:
  name: Hash merge behaviour
  default: replace
  type: string
  choices: ["replace", "merge"]
  description:
    - This setting controls how variables merge in Ansible.
      By default Ansible will override variables in specific precedence orders, as described in Variables.
      When a variable of higher precedence wins, it will replace the other value.
    - "Some users prefer that variables that are hashes (aka 'dictionaries' in Python terms) are merged.
      This setting is called 'merge'. This is not the default behavior and it does not affect variables whose values are scalars
      (integers, strings) or arrays.  We generally recommend not using this setting unless you think you have an absolute need for it,
      and playbooks in the official examples repos do not use this setting"
    - In version 2.0 a ``combine`` filter was added to allow doing this for a particular variable (described in Filters).
  env: [{name: ANSIBLE_HASH_BEHAVIOUR}]
  ini:
  - {key: hash_behaviour, section: defaults}
DEFAULT_HOST_LIST:
  name: Inventory Source
  default: /etc/ansible/hosts
  description: Colon separated list of Ansible inventory sources
  env:
    - name: ANSIBLE_HOSTS
      deprecated:
        why: The variable is misleading as it can be a list of hosts and/or paths to inventory sources
        version: "2.8"
        alternatives: ANSIBLE_INVENTORY
    - name: ANSIBLE_INVENTORY
  expand_relative_paths: True
  ini:
    - key: hostfile
      section: defaults
      deprecated:
        why: The key is misleading as it can also be a list of hosts, a directory or a list of paths
        version: "2.8"
        alternatives: "[defaults]\ninventory=/path/to/file|dir"
    - key: inventory
      section: defaults
  type: pathlist
  yaml: {key: defaults.inventory}
DEFAULT_INTERNAL_POLL_INTERVAL:
  name: Internal poll interval
  default: 0.001
  env: []
  ini:
  - {key: internal_poll_interval, section: defaults}
  type: float
  version_added: "2.2"
  description:
    - This sets the interval (in seconds) of Ansible internal processes polling each other.
      Lower values improve performance with large playbooks at the expense of extra CPU load.
      Higher values are more suitable for Ansible usage in automation scenarios,
      when UI responsiveness is not required but CPU usage might be a concern.
    - "The default corresponds to the value hardcoded in Ansible <= 2.1"
DEFAULT_INVENTORY_PLUGIN_PATH:
  name: Inventory Plugins Path
  default: ~/.ansible/plugins/inventory:/usr/share/ansible/plugins/inventory
  description: Colon separated paths in which Ansible will search for Inventory Plugins.
  env: [{name: ANSIBLE_INVENTORY_PLUGINS}]
  ini:
  - {key: inventory_plugins, section: defaults}
  type: pathspec
DEFAULT_JINJA2_EXTENSIONS:
  name: Enabled Jinja2 extensions
  default: []
  description:
    - This is a developer-specific feature that allows enabling additional Jinja2 extensions.
    - "See the Jinja2 documentation for details. If you do not know what these do, you probably don't need to change this setting :)"
  env: [{name: ANSIBLE_JINJA2_EXTENSIONS}]
  ini:
  - {key: jinja2_extensions, section: defaults}
DEFAULT_KEEP_REMOTE_FILES:
  name: Keep remote files
  default: False
  description: Enables/disables the cleaning up of the temporary files Ansible used to execute the tasks on the remote.
  env: [{name: ANSIBLE_KEEP_REMOTE_FILES}]
  ini:
  - {key: keep_remote_files, section: defaults}
  type: boolean
DEFAULT_LIBVIRT_LXC_NOSECLABEL:
  # TODO: move to plugin
  name: No security label on Lxc
  default: False
  description:
    - "This setting causes libvirt to connect to lxc containers by passing --noseclabel to virsh.
      This is necessary when running on systems which do not have SELinux."
  env: [{name: LIBVIRT_LXC_NOSECLABEL}]
  ini:
  - {key: libvirt_lxc_noseclabel, section: selinux}
  type: boolean
  version_added: "2.1"
DEFAULT_LOAD_CALLBACK_PLUGINS:
  name: Load callbacks for adhoc
  default: False
  description:
    - Controls whether callback plugins are loaded when running /usr/bin/ansible.
      This may be used to log activity from the command line, send notifications, and so on.
      Callback plugins are always loaded for ``ansible-playbook``.
  env: [{name: ANSIBLE_LOAD_CALLBACK_PLUGINS}]
  ini:
  - {key: bin_ansible_callbacks, section: defaults}
  type: boolean
  version_added: "1.8"
DEFAULT_LOCAL_TMP:
  name: Controller temporary directory
  default: ~/.ansible/tmp
  description: Temporary directory for Ansible to use on the controller.
  env: [{name: ANSIBLE_LOCAL_TEMP}]
  ini:
  - {key: local_tmp, section: defaults}
  type: tmppath
DEFAULT_LOG_PATH:
  name: Ansible log file path
  default: ''
  description: File to which Ansible will log on the controller. When empty logging is disabled.
  env: [{name: ANSIBLE_LOG_PATH}]
  ini:
  - {key: log_path, section: defaults}
  type: path
DEFAULT_LOOKUP_PLUGIN_PATH:
  name: Lookup Plugins Path
  description: Colon separated paths in which Ansible will search for Lookup Plugins.
  default: ~/.ansible/plugins/lookup:/usr/share/ansible/plugins/lookup
  env: [{name: ANSIBLE_LOOKUP_PLUGINS}]
  ini:
  - {key: lookup_plugins, section: defaults}
  type: pathspec
  yaml: {key: defaults.lookup_plugins}
DEFAULT_MANAGED_STR:
  name: Ansible managed
  default: 'Ansible managed'
  description: Sets the macro for the 'ansible_managed' variable available for M(template) tasks.
  env: []
  ini:
  - {key: ansible_managed, section: defaults}
  yaml: {key: defaults.ansible_managed}
DEFAULT_MODULE_ARGS:
  name: Adhoc default arguments
  default: ''
  description:
    - This sets the default arguments to pass to the ``ansible`` adhoc binary if no ``-a`` is specified.
  env: [{name: ANSIBLE_MODULE_ARGS}]
  ini:
  - {key: module_args, section: defaults}
DEFAULT_MODULE_COMPRESSION:
  name: Python module compression
  default: ZIP_DEFLATED
  description: Compression scheme to use when transfering Python modules to the target.
  env: []
  ini:
  - {key: module_compression, section: defaults}
# vars:
#   - name: ansible_module_compression
DEFAULT_MODULE_LANG:
  name: Target language environment
  default: "{{CONTROLER_LANG}}"
  description: "Language locale setting to use for modules when they execute on the target, if empty it defaults to 'en_US.UTF-8'"
  env: [{name: ANSIBLE_MODULE_LANG}]
  ini:
  - {key: module_lang, section: defaults}
# vars:
#   - name: ansible_module_lang
DEFAULT_MODULE_NAME:
  name: Default adhoc module
  default: command
  description: "Module to use with the ``ansible`` AdHoc command, if none is specified via ``-m``."
  env: []
  ini:
  - {key: module_name, section: defaults}
DEFAULT_MODULE_PATH:
  name: Modules Path
  description: Colon separated paths in which Ansible will search for Modules.
  default: ~/.ansible/plugins/modules:/usr/share/ansible/plugins/modules
  env: [{name: ANSIBLE_LIBRARY}]
  ini:
  - {key: library, section: defaults}
  type: pathspec
DEFAULT_MODULE_SET_LOCALE:
  name: Target locale
  default: False
  description: Controls if we set locale for modules when executing on the target.
  env: [{name: ANSIBLE_MODULE_SET_LOCALE}]
  ini:
  - {key: module_set_locale, section: defaults}
  type: boolean
# vars:
#   - name: ansible_module_locale
DEFAULT_MODULE_UTILS_PATH:
  name: Module Utils Path
  description: Colon separated paths in which Ansible will search for Module utils files, which are shared by modules.
  default: ~/.ansible/plugins/module_utils:/usr/share/ansible/plugins/module_utils
  env: [{name: ANSIBLE_MODULE_UTILS}]
  ini:
  - {key: module_utils, section: defaults}
  type: pathspec
DEFAULT_NO_LOG:
  name: No log
  default: False
  description: "Toggle Ansible's display and logging of task details, mainly used to avoid security disclosures."
  env: [{name: ANSIBLE_NO_LOG}]
  ini:
  - {key: no_log, section: defaults}
  type: boolean
DEFAULT_NO_TARGET_SYSLOG:
  name: No syslog on target
  default: False
  description: Toggle Ansbile logging to syslog on the target when it executes tasks.
  env: [{name: ANSIBLE_NO_TARGET_SYSLOG}]
  ini:
  - {key: no_target_syslog, section: defaults}
  type: boolean
  yaml: {key: defaults.no_target_syslog}
DEFAULT_NULL_REPRESENTATION:
  name: Represent a null
  default: ~
  description: What templating should return as a 'null' value. When not set it will let Jinja2 decide.
  env: [{name: ANSIBLE_NULL_REPRESENTATION}]
  ini:
  - {key: null_representation, section: defaults}
  type: none
DEFAULT_POLL_INTERVAL:
  name: Async poll interval
  default: 15
  description:
    - For asynchronous tasks in Ansible (covered in Asynchronous Actions and Polling),
      this is how often to check back on the status of those tasks when an explicit poll interval is not supplied.
      The default is a reasonably moderate 15 seconds which is a tradeoff between checking in frequently and
      providing a quick turnaround when something may have completed.
  env: [{name: ANSIBLE_POLL_INTERVAL}]
  ini:
  - {key: poll_interval, section: defaults}
  type: integer
DEFAULT_PRIVATE_KEY_FILE:
  name: Private key file
  default: ~
  description:
    - Option for connections using a certificate or key file to authenticate, rather than an agent or passwords,
      you can set the default value here to avoid re-specifying --private-key with every invocation.
  env: [{name: ANSIBLE_PRIVATE_KEY_FILE}]
  ini:
  - {key: private_key_file, section: defaults}
  type: path
DEFAULT_PRIVATE_ROLE_VARS:
  name: Private role variables
  default: False
  description: ''
  env: [{name: ANSIBLE_PRIVATE_ROLE_VARS}]
  ini:
  - {key: private_role_vars, section: defaults}
  type: boolean
  yaml: {key: defaults.private_role_vars}
DEFAULT_REMOTE_PORT:
  name: Remote port
  default: ~
  description: Port to use in remote connections, when blank it will use the connection plugin default.
  env: [{name: ANSIBLE_REMOTE_PORT}]
  ini:
  - {key: remote_port, section: defaults}
  type: integer
  yaml: {key: defaults.remote_port}
DEFAULT_REMOTE_TMP:
  name: Target temporary directory
  default: ~/.ansible/tmp
  description:
    - Temporary directory to use on targets when executing tasks.
    - In some cases Ansible may still choose to use a system temporary dir to avoid permission issues.
  env: [{name: ANSIBLE_REMOTE_TEMP}]
  ini:
  - {key: remote_tmp, section: defaults}
  vars:
    - name: ansible_remote_tmp
DEFAULT_REMOTE_USER:
  name: Login/Remote User
  default:
  description:
    - Sets the login user for the target machines
    - "When blank it uses the connection plugin's default, normally the user currently executing Ansible."
  env: [{name: ANSIBLE_REMOTE_USER}]
  ini:
  - {key: remote_user, section: defaults}
DEFAULT_ROLES_PATH:
  name: Roles path
  default: ~/.ansible/roles:/usr/share/ansible/roles:/etc/ansible/roles
  description: Colon separated paths in which Ansible will search for Roles.
  env: [{name: ANSIBLE_ROLES_PATH}]
  expand_relative_paths: True
  ini:
  - {key: roles_path, section: defaults}
  type: pathspec
  yaml: {key: defaults.roles_path}
DEFAULT_SCP_IF_SSH:
  # TODO: move to ssh plugin
  default: smart
  description:
    - "Prefered method to use when transfering files over ssh"
    - When set to smart, Ansible will try them until one succeeds or they all fail
    - If set to True, it will force 'scp', if False it will use 'sftp'
  env: [{name: ANSIBLE_SCP_IF_SSH}]
  ini:
  - {key: scp_if_ssh, section: ssh_connection}
DEFAULT_SELINUX_SPECIAL_FS:
  name: Problematic file systems
  default: fuse, nfs, vboxsf, ramfs, 9p
  description:
    - "Some filesystems do not support safe operations and/or return inconsistent errors,
       this setting makes Ansible 'tolerate' those in the list w/o causing fatal errors."
    - Data corruption may occur and writes are not always verified when a filesystem is in the list.
  env: []
  ini:
  - {key: special_context_filesystems, section: selinux}
  type: list
DEFAULT_SFTP_BATCH_MODE:
  # TODO: move to ssh plugin
  default: True
  description: 'TODO: write it'
  env: [{name: ANSIBLE_SFTP_BATCH_MODE}]
  ini:
  - {key: sftp_batch_mode, section: ssh_connection}
  type: boolean
  yaml: {key: ssh_connection.sftp_batch_mode}
DEFAULT_SQUASH_ACTIONS:
  name: Squashable actions
  default: apk, apt, dnf, homebrew, openbsd_pkg, pacman, pkgng, yum, zypper
  description:
    - Ansible can optimise actions that call modules that support list parameters when using ``with_`` looping.
      Instead of calling the module once for each item, the module is called once with the full list.
    - The default value for this setting is only for certain package managers, but it can be used for any module
    - Currently, this is only supported for modules that have a name or pkg parameter, and only when the item is the only thing being passed to the parameter.
  env: [{name: ANSIBLE_SQUASH_ACTIONS}]
  ini:
  - {key: squash_actions, section: defaults}
  type: list
  version_added: "2.0"
DEFAULT_SSH_TRANSFER_METHOD:
  # TODO: move to ssh plugin
  default:
  description: 'unused?'
  #  - "Prefered method to use when transfering files over ssh"
  #  - Setting to smart will try them until one succeeds or they all fail
  #choices: ['sftp', 'scp', 'dd', 'smart']
  env: [{name: ANSIBLE_SSH_TRANSFER_METHOD}]
  ini:
  - {key: transfer_method, section: ssh_connection}
DEFAULT_STDOUT_CALLBACK:
  name: Main display callback plugin
  default: default
  description:
    - "Set the main callback used to display Ansible output, you can only have one at a time."
    - You can have many other callbacks, but just one can be in charge of stdout.
  env: [{name: ANSIBLE_STDOUT_CALLBACK}]
  ini:
  - {key: stdout_callback, section: defaults}
DEFAULT_STRATEGY:
  name: Implied strategy
  default: 'linear'
  description: Set the default strategy used for plays.
  env: [{name: ANSIBLE_STRATEGY}]
  ini:
  - {key: strategy, section: defaults}
  version_added: "2.3"
DEFAULT_STRATEGY_PLUGIN_PATH:
  name: Strategy Plugins Path
  description: Colon separated paths in which Ansible will search for Strategy Plugins.
  default: ~/.ansible/plugins/strategy:/usr/share/ansible/plugins/strategy
  env: [{name: ANSIBLE_STRATEGY_PLUGINS}]
  ini:
  - {key: strategy_plugins, section: defaults}
  type: pathspec
DEFAULT_SU:
  default: False
  description: 'Toggle the use of "su" for tasks.'
  env: [{name: ANSIBLE_SU}]
  ini:
  - {key: su, section: defaults}
  type: boolean
  yaml: {key: defaults.su}
DEFAULT_SUDO:
  default: False
  deprecated:
    why: In favor of become which is a generic framework
    version: "2.8"
    alternatives: become
  description: 'Toggle the use of "sudo" for tasks.'
  env: [{name: ANSIBLE_SUDO}]
  ini:
  - {key: sudo, section: defaults}
  type: boolean
DEFAULT_SUDO_EXE:
  name: sudo executable
  default: sudo
  deprecated:
    why: In favor of become which is a generic framework
    version: "2.8"
    alternatives: become
  description: 'specify an "sudo" executable, otherwise it relies on PATH.'
  env: [{name: ANSIBLE_SUDO_EXE}]
  ini:
  - {key: sudo_exe, section: defaults}
DEFAULT_SUDO_FLAGS:
  name: sudo flags
  default: '-H -S -n'
  deprecated:
    why: In favor of become which is a generic framework
    version: "2.8"
    alternatives: become
  description: 'Flags to pass to "sudo"'
  env: [{name: ANSIBLE_SUDO_FLAGS}]
  ini:
  - {key: sudo_flags, section: defaults}
DEFAULT_SUDO_USER:
  name: sudo user
  default:
  deprecated:
    why: In favor of become which is a generic framework
    version: "2.8"
    alternatives: become
  description: 'User you become when using "sudo", leaving it blank will use the default configured on the target (normally root)'
  env: [{name: ANSIBLE_SUDO_USER}]
  ini:
  - {key: sudo_user, section: defaults}
DEFAULT_SU_EXE:
  name: su executable
  default: su
  deprecated:
    why: In favor of become which is a generic framework
    version: "2.8"
    alternatives: become
  description: 'specify an "su" executable, otherwise it relies on PATH.'
  env: [{name: ANSIBLE_SU_EXE}]
  ini:
  - {key: su_exe, section: defaults}
DEFAULT_SU_FLAGS:
  name: su flags
  default: ''
  deprecated:
    why: In favor of become which is a generic framework
    version: "2.8"
    alternatives: become
  description: 'Flags to pass to su'
  env: [{name: ANSIBLE_SU_FLAGS}]
  ini:
  - {key: su_flags, section: defaults}
DEFAULT_SU_USER:
  name: su user
  default:
  description: 'User you become when using "su", leaving it blank will use the default configured on the target (normally root)'
  env: [{name: ANSIBLE_SU_USER}]
  ini:
  - {key: su_user, section: defaults}
  deprecated:
    why: In favor of become which is a generic framework
    version: "2.8"
    alternatives: become
DEFAULT_SYSLOG_FACILITY:
  name: syslog facility
  default: LOG_USER
  description: Syslog facility to use when Ansible logs to the remote target
  env: [{name: ANSIBLE_SYSLOG_FACILITY}]
  ini:
  - {key: syslog_facility, section: defaults}
DEFAULT_TASK_INCLUDES_STATIC:
  name: Task include static
  default: False
  description:
    - The `include` tasks can be static or dynamic, this toggles the default expected behaviour if autodetection fails and it is not explicitly set in task.
  env: [{name: ANSIBLE_TASK_INCLUDES_STATIC}]
  ini:
  - {key: task_includes_static, section: defaults}
  type: boolean
  version_added: "2.1"
  deprecated:
    why: include itself is deprecated and this setting will not matter in the future
    version: "2.8"
    alternatives: None, as its already built into the decision between include_tasks and import_tasks
DEFAULT_TEST_PLUGIN_PATH:
  name: Jinja2 Test Plugins Path
  description: Colon separated paths in which Ansible will search for Jinja2 Test Plugins.
  default: ~/.ansible/plugins/test:/usr/share/ansible/plugins/test
  env: [{name: ANSIBLE_TEST_PLUGINS}]
  ini:
  - {key: test_plugins, section: defaults}
  type: pathspec
DEFAULT_TIMEOUT:
  name: Connection timeout
  default: 10
  description: This is the default timeout for connection plugins to use.
  env: [{name: ANSIBLE_TIMEOUT}]
  ini:
  - {key: timeout, section: defaults}
  type: integer
DEFAULT_TRANSPORT:
  name: Connection plugin
  default: smart
  description: "Default connection plugin to use, the 'smart' option will toggle between 'ssh' and 'paramiko' depending on controller OS and ssh versions"
  env: [{name: ANSIBLE_TRANSPORT}]
  ini:
  - {key: transport, section: defaults}
DEFAULT_UNDEFINED_VAR_BEHAVIOR:
  name: Jinja2 fail on undefined
  default: True
  version_added: "1.3"
  description:
    - When True, this causes ansible templating to fail steps that reference variable names that are likely typoed.
    - "Otherwise, any '{{ template_expression }}' that contains undefined variables will be rendered in a template or ansible action line exactly as written."
  env: [{name: ANSIBLE_ERROR_ON_UNDEFINED_VARS}]
  ini:
  - {key: error_on_undefined_vars, section: defaults}
  type: boolean
DEFAULT_VARS_PLUGIN_PATH:
  name: Vars Plugins Path
  default: ~/.ansible/plugins/vars:/usr/share/ansible/plugins/vars
  description: Colon separated paths in which Ansible will search for Vars Plugins.
  env: [{name: ANSIBLE_VARS_PLUGINS}]
  ini:
  - {key: vars_plugins, section: defaults}
  type: pathspec
# TODO: unused?
#DEFAULT_VAR_COMPRESSION_LEVEL:
#  default: 0
#  description: 'TODO: write it'
#  env: [{name: ANSIBLE_VAR_COMPRESSION_LEVEL}]
#  ini:
#  - {key: var_compression_level, section: defaults}
#  type: integer
#  yaml: {key: defaults.var_compression_level}
DEFAULT_VAULT_ID_MATCH:
  name: Force vault id match
  default: False
  description: 'If true, decrypting vaults with a vault id will only try the password from the matching vault-id'
  env: [{name: ANSIBLE_VAULT_ID_MATCH}]
  ini:
  - {key: vault_id_match, section: defaults}
  yaml: {key: defaults.vault_id_match}
DEFAULT_VAULT_IDENTITY:
  name: Vault id label
  default: default
  description: 'The label to use for the default vault id label in cases where a vault id label is not provided'
  env: [{name: ANSIBLE_VAULT_IDENTITY}]
  ini:
  - {key: vault_identity, section: defaults}
  yaml: {key: defaults.vault_identity}
DEFAULT_VAULT_IDENTITY_LIST:
  name: Default vault ids
  default: []
  description: 'A list of vault-ids to use by default. Equivalent to multiple --vault-id args. Vault-ids are tried in order.'
  env: [{name: ANSIBLE_VAULT_IDENTITY_LIST}]
  ini:
  - {key: vault_identity_list, section: defaults}
  type: list
  yaml: {key: defaults.vault_identity_list}
DEFAULT_VAULT_PASSWORD_FILE:
  name: Vault password file
  default: ~
  description: 'The vault password file to use. Equivalent to --vault-password-file or --vault-id'
  env: [{name: ANSIBLE_VAULT_PASSWORD_FILE}]
  ini:
  - {key: vault_password_file, section: defaults}
  type: path
  yaml: {key: defaults.vault_password_file}
DEFAULT_VERBOSITY:
  name: Verbosity
  default: 0
  description: Sets the default verbosity, equivalent to the number of ``-v`` passed in the command line.
  env: [{name: ANSIBLE_VERBOSITY}]
  ini:
  - {key: verbosity, section: defaults}
  type: integer
DEPRECATION_WARNINGS:
  name: Deprecation messages
  default: True
  description: "Toggle to control the showing of deprecation warnings"
  env: [{name: ANSIBLE_DEPRECATION_WARNINGS}]
  ini:
  - {key: deprecation_warnings, section: defaults}
  type: boolean
DIFF_ALWAYS:
  name: Show differences
  default: False
  description: Configuration toggle to tell modules to show differences when in 'changed' status, equivalent to ``--diff``.
  env: [{name: ANSIBLE_DIFF_ALWAYS}]
  ini:
  - {key: always, section: diff}
  type: bool
DIFF_CONTEXT:
  name: Difference context
  default: 3
  description: How many lines of context to show when displaying the differences between files.
  env: [{name: ANSIBLE_DIFF_CONTEXT}]
  ini:
  - {key: context, section: diff}
  type: integer
DISPLAY_ARGS_TO_STDOUT:
  name: Show task arguments
  default: False
  description:
    - "Normally ``ansible-playbook`` will print a header for each task that is run.
      These headers will contain the name: field from the task if you specified one.
      If you didn't then ``ansible-playbook`` uses the task's action to help you tell which task is presently running.
      Sometimes you run many of the same action and so you want more information about the task to differentiate it from others of the same action.
      If you set this variable to True in the config then ``ansible-playbook`` will also include the task's arguments in the header."
    - "This setting defaults to False because there is a chance that you have sensitive values in your parameters and
      you do not want those to be printed."
    - "If you set this to True you should be sure that you have secured your environment's stdout
      (no one can shoulder surf your screen and you aren't saving stdout to an insecure file) or
      made sure that all of your playbooks explicitly added the ``no_log: True`` parameter to tasks which have sensistive values
      See How do I keep secret data in my playbook? for more information."
  env: [{name: ANSIBLE_DISPLAY_ARGS_TO_STDOUT}]
  ini:
  - {key: display_args_to_stdout, section: defaults}
  type: boolean
  version_added: "2.1"
DISPLAY_SKIPPED_HOSTS:
  name: Show skipped results
  default: True
  description: "Toggle to control displaying skipped task/host entries in a task in the default callback"
  env: [{name: DISPLAY_SKIPPED_HOSTS}]
  ini:
  - {key: display_skipped_hosts, section: defaults}
  type: boolean
ERROR_ON_MISSING_HANDLER:
  name: Missing handler error
  default: True
  description: "Toggle to allow missing handlers to become a warning instead of an error when notifying."
  env: [{name: ANSIBLE_ERROR_ON_MISSING_HANDLER}]
  ini:
  - {key: error_on_missing_handler, section: defaults}
  type: boolean
GALAXY_IGNORE_CERTS:
  name: Galaxy validate certs
  default: False
  description:
    - If set to yes, ansible-galaxy will not validate TLS certificates.
      This can be useful for testing against a server with a self-signed certificate.
  env: [{name: ANSIBLE_GALAXY_IGNORE}]
  ini:
  - {key: ignore_certs, section: galaxy}
  type: boolean
GALAXY_ROLE_SKELETON:
  name: Galaxy skeleton direcotry
  default:
  description: Role skeleton directory to use as a template for the ``init`` action in ``ansible-galaxy``, same as ``--role-skeleton``.
  env: [{name: ANSIBLE_GALAXY_ROLE_SKELETON}]
  ini:
  - {key: role_skeleton, section: galaxy}
  type: path
GALAXY_ROLE_SKELETON_IGNORE:
  name: Galaxy skeleton ignore
  default: ["^.git$", "^.*/.git_keep$"]
  description: patterns of files to ignore inside a galaxy role skeleton directory
  env: [{name: ANSIBLE_GALAXY_ROLE_SKELETON_IGNORE}]
  ini:
  - {key: role_skeleton_ignore, section: galaxy}
  type: list
# TODO: unused?
#GALAXY_SCMS:
#  name: Galaxy SCMS
#  default: git, hg
#  description: Available galaxy source control management systems.
#  env: [{name: ANSIBLE_GALAXY_SCMS}]
#  ini:
#  - {key: scms, section: galaxy}
#  type: list
GALAXY_SERVER:
  default: https://galaxy.ansible.com
  description: "URL to prepend when roles don't specify the full URI, assume they are referencing this server as the source."
  env: [{name: ANSIBLE_GALAXY_SERVER}]
  ini:
  - {key: server, section: galaxy}
  yaml: {key: galaxy.server}
HOST_KEY_CHECKING:
  name: Check host keys
  default: True
  description: 'Set this to "False" if you want to avoid host key checking by the underlying tools Ansible uses to connect to the host'
  env: [{name: ANSIBLE_HOST_KEY_CHECKING}]
  ini:
  - {key: host_key_checking, section: defaults}
  type: boolean
INVENTORY_ENABLED:
  name: Active Inventory plugins
  default: ['host_list', 'script', 'yaml', 'ini']
  description: List of enabled inventory plugins, it also determines the order in which they are used.
  env: [{name: ANSIBLE_INVENTORY_ENABLED}]
  ini:
  - {key: enable_plugins, section: inventory}
  type: list
INVENTORY_IGNORE_EXTS:
  name: Inventory ignore extensions
  default: "{{(BLACKLIST_EXTS + ( '~', '.orig', '.ini', '.cfg', '.retry'))}}"
  description: List of extensions to ignore when using a directory as an inventory source
  env: [{name: ANSIBLE_INVENTORY_IGNORE}]
  ini:
  - {key: inventory_ignore_extensions, section: defaults}
  - {key: ignore_extensions, section: inventory}
  type: list
INVENTORY_IGNORE_PATTERNS:
  name: Inventory ignore patterns
  default: []
  description: List of patterns to ignore when using a directory as an inventory source
  env: [{name: ANSIBLE_INVENTORY_IGNORE_REGEX}]
  ini:
  - {key: inventory_ignore_patterns, section: defaults}
  - {key: ignore_patterns, section: inventory}
  type: list
INVENTORY_UNPARSED_IS_FAILED:
  name: Unparsed Inventory failure
  default: False
  description: If 'true' unparsed inventory sources become fatal errors, they are warnings otherwise.
  env: [{name: ANSIBLE_INVENTORY_UNPARSED_FAILED}]
  ini:
  - {key: unparsed_is_failed, section: inventory}
  type: bool
MAX_FILE_SIZE_FOR_DIFF:
  name: Diff maxiumum file size
  default: 104448
  description: Maximum size of files to be considered for diff display
  env: [{name: ANSIBLE_MAX_DIFF_SIZE}]
  ini:
  - {key: max_diff_size, section: defaults}
  type: int
MERGE_MULTIPLE_CLI_TAGS:
  name: Merge 'tags' options
  default: True
  description:
    - "This allows changing how multiple --tags and --skip-tags arguments are handled on the command line.
      In Ansible up to and including 2.3, specifying --tags more than once will only take the last value of --tags."
    - "Setting this config value to True will mean that all of the --tags options will be merged together. The same holds true for --skip-tags."
  env: [{name: ANSIBLE_MERGE_MULTIPLE_CLI_TAGS}]
  ini:
    - {key: merge_multiple_cli_tags, section: defaults}
  type: bool
  version_added: "2.3"
NETWORK_GROUP_MODULES:
  name: Network module families
  default: [eos, nxos, ios, iosxr, junos, ce, vyos, sros, dellos9, dellos10, dellos6, asa, aruba, aireos]
  description: 'TODO: write it'
  env: [{name: NETWORK_GROUP_MODULES}]
  ini:
  - {key: network_group_modules, section: defaults}
  type: list
  yaml: {key: defaults.network_group_modules}
#ONLY_NAMESPACE_FACTS:
# Deffered to 2.5
# FIXME: reenable when we can remove ansible_ prefix from namespaced facts
#  default: False
#  description:
#    - Facts normally get injected as top level variables, this setting prevents that.
#    - Facts are still available in the `ansible_facts` variable w/o the `ansible_` prefix.
#  env: [{name: ANSIBLE_RESTRICT_FACTS}]
#  ini:
#  - {key: restrict_facts_namespace, section: defaults}
#  type: boolean
#  yaml: {key: defaults.restrict_facts_namespace}
#  version_added: "2.4"
PARAMIKO_HOST_KEY_AUTO_ADD:
  # TODO: move to plugin
  default: False
  description: 'TODO: write it'
  env: [{name: ANSIBLE_PARAMIKO_HOST_KEY_AUTO_ADD}]
  ini:
  - {key: host_key_auto_add, section: paramiko_connection}
  type: boolean
PARAMIKO_LOOK_FOR_KEYS:
  # TODO: move to plugin
  default: True
  description: 'TODO: write it'
  env: [{name: ANSIBLE_PARAMIKO_LOOK_FOR_KEYS}]
  ini:
  - {key: look_for_keys, section: paramiko_connection}
  type: boolean
PARAMIKO_PROXY_COMMAND:
  # TODO: move to plugin
  default:
  description: 'TODO: write it'
  env: [{name: ANSIBLE_PARAMIKO_PROXY_COMMAND}]
  ini:
  - {key: proxy_command, section: paramiko_connection}
PARAMIKO_PTY:
  # TODO: move to plugin
  default: True
  description: 'TODO: write it'
  env: [{name: ANSIBLE_PARAMIKO_PTY}]
  ini:
  - {key: pty, section: paramiko_connection}
  type: boolean
PARAMIKO_RECORD_HOST_KEYS:
  # TODO: move to plugin
  default: True
  description: 'TODO: write it'
  env: [{name: ANSIBLE_PARAMIKO_RECORD_HOST_KEYS}]
  ini:
  - {key: record_host_keys, section: paramiko_connection}
  type: boolean
PERSISTENT_CONTROL_PATH_DIR:
  name: Persistence socket path
  default: ~/.ansible/pc
  description: Path to socket to be used by the connection persistence system.
  env: [{name: ANSIBLE_PERSISTENT_CONTROL_PATH_DIR}]
  ini:
  - {key: control_path_dir, section: persistent_connection}
  type: path
PERSISTENT_CONNECT_TIMEOUT:
  name: Persistence timeout
  default: 30
  description: This controls how long the persistent connection will remain idle before it is destroyed.
  env: [{name: ANSIBLE_PERSISTENT_CONNECT_TIMEOUT}]
  ini:
  - {key: connect_timeout, section: persistent_connection}
  type: integer
PERSISTENT_CONNECT_RETRY_TIMEOUT:
  name: Persistence connection retry timeout
  default: 15
  description: This contorls the retry timeout for presistent connection to connect to the local domain socket.
  env: [{name: ANSIBLE_PERSISTENT_CONNECT_RETRY_TIMEOUT}]
  ini:
  - {key: connect_retry_timeout, section: persistent_connection}
  type: integer
PERSISTENT_COMMAND_TIMEOUT:
  name: Persistence command timeout
  default: 10
  description: This controls the amount of time to wait for response from remote device before timing out presistent connection.
  env: [{name: ANSIBLE_PERSISTENT_COMMAND_TIMEOUT}]
  ini:
  - {key: command_timeout, section: persistent_connection}
  type: int
PLAYBOOK_VARS_ROOT:
  name: playbook vars files root
  default: top
  version_added: "2.4.1"
  description:
    - This sets which playbook dirs will be used as a root to process vars plugins, which includes finding host_vars/group_vars
    - The ``top`` option follows the traditional behaviour of using the top playbook in the chain to find the root directory.
    - The ``bottom`` option follows the 2.4.0 behaviour of using the current playbook to find the root directory.
    - The ``all`` option examines from the first parent to the current playbook.
  env: [{name: ANSIBLE_PLAYBOOK_VARS_ROOT}]
  ini:
  - {key: playbook_vars_root, section: defaults}
  choices: [ top, bottom, all ]
RETRY_FILES_ENABLED:
  name: Retry files
  default: True
  description: This controls whether a failed Ansible playbook should create a .retry file.
  env: [{name: ANSIBLE_RETRY_FILES_ENABLED}]
  ini:
  - {key: retry_files_enabled, section: defaults}
  type: bool
RETRY_FILES_SAVE_PATH:
  name: Retry files path
  default: ~
  description: This sets the path in which Ansible will save .retry files when a playbook fails and retry files are enabled.
  env: [{name: ANSIBLE_RETRY_FILES_SAVE_PATH}]
  ini:
  - {key: retry_files_save_path, section: defaults}
  type: path
SHOW_CUSTOM_STATS:
  name: Display custom stats
  default: False
  description: 'This adds the custom stats set via the set_stats plugin to the default output'
  env: [{name: ANSIBLE_SHOW_CUSTOM_STATS}]
  ini:
  - {key: show_custom_stats, section: defaults}
  type: bool
STRING_TYPE_FILTERS:
  name: Filters to preserve strings
  default: [string, to_json, to_nice_json, to_yaml, ppretty, json]
  description:
    - "This list of filters avoids 'type conversion' when templating variables"
    - Useful when you want to avoid conversion into lists or dictionaries for JSON strings, for example.
  env: [{name: ANSIBLE_STRING_TYPE_FILTERS}]
  ini:
  - {key: dont_type_filters, section: jinja2}
  type: list
SYSTEM_WARNINGS:
  name: System warnings
  default: True
  description:
    - Allows disabling of warnings related to potential issues on the system running ansible itself (not on the managed hosts)
    - These may include warnings about 3rd party packages or other conditions that should be resolved if possible.
  env: [{name: ANSIBLE_SYSTEM_WARNINGS}]
  ini:
  - {key: system_warnings, section: defaults}
  type: boolean
USE_PERSISTENT_CONNECTIONS:
  name: Persistence
  default: False
  description: Toggles the use of persistence for connections.
  env: [{name: ANSIBLE_USE_PERSISTENT_CONNECTIONS}]
  ini:
  - {key: use_persistent_connections, section: defaults}
  type: boolean
VARIABLE_PRECEDENCE:
  name: Group variable precedence
  default: ['all_inventory', 'groups_inventory', 'all_plugins_inventory', 'all_plugins_play', 'groups_plugins_inventory', 'groups_plugins_play']
  description: Allows to change the group variable precedence merge order.
  env: [{name: ANSIBLE_PRECEDENCE}]
  ini:
  - {key: precedence, section: defaults}
  type: list
  version_added: "2.4"
YAML_FILENAME_EXTENSIONS:
  name: Valid YAML extensions
  default: [".yml", ".yaml", ".json"]
  description:
    - "Check all of these extensions when looking for 'variable' files which should be YAML or JSON or vaulted versions of these."
    - 'This affects vars_files, include_vars, inventory and vars plugins among others.'
  env:
    - name: ANSIBLE_YAML_FILENAME_EXT
  ini:
    - section: defaults
      key: yaml_valid_extensions
  type: list
...

こいつらを使うとこんな感じになります。イメージ通り。

$ ansible-playbook -i hosts echo-sd.yml

_人人人人人人人人人人_
> PLAY [localhost] <
 ̄Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^ ̄


_人人人人人人人人人人人人人_
> TASK [Gathering Facts] <
 ̄Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^ ̄

ok: [localhost]

_人人人人人人人人人人人人人_
> TASK [Display message] <
 ̄Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^ ̄

ok: [localhost] => {
    "msg": "突然の死"
}

_人人人人人人人_
> PLAY RECAP <
 ̄Y^Y^Y^Y^Y^Y^Y^ ̄

localhost                  : ok=2    changed=0    unreachable=0    failed=0

cowsayバージョンはシュールな感じがしましたが、echo-sd版はなんというか...うるさいですね。出力がやかましい。
当初の目的は達成したものの、特に喜びがなくて困りました。

解説

それではecho-sd化に関係する部分について解説したいと思います。

処理の流れ

display.py では、タスク実行時のメッセージを画面を出力しますが、cowsayに関しては概ね以下のような流れで読み込み、実行されます。

  1. 読み込み
    • cowsayパスの決定
    • cowsayキャラクタの除外
  2. 実行
    • cowsayバナー出力

それぞれの処理を確認し、echo-sd化に向けて置き換えていきましょう。

読み込み

Ansible起動時には各種モジュールが読み込まれますが、その流れの中でDisplayが読み込まれ、 init 処理が実行されます。

init
    def __init__(self, verbosity=0):

        ()

        # cowsayの実行パスを格納する変数
        self.b_cowsay = None

        # cowsayキャラクタをコンフィグから読み込む
        self.noncow = C.ANSIBLE_COW_SELECTION

        # cowsayの実行パスを決定
        self.set_cowsay_info()

        # コンフィグ設定と合わせて使用可能なcowsayキャラクタを決定
        if self.b_cowsay:
            try:
                # cowsay -lコマンドから環境上使用可能なcowsayキャラクタを取得
                cmd = subprocess.Popen([self.b_cowsay, "-l"], stdout=subprocess.PIPE, stderr=subprocess.PIPE)
                (out, err) = cmd.communicate()
                self.cows_available = set([to_text(c) for c in out.split()])

                # 使用可能なもののうち、ホワイトリストでも指定されたキャラクタのみを抜き出す
                if C.ANSIBLE_COW_WHITELIST:
                    self.cows_available = set(C.ANSIBLE_COW_WHITELIST).intersection(self.cows_available)
            except:
                # cowsay -lコマンドが実行できないなど、何らかの例外が起きたらcowsayしないようにする
                self.b_cowsay = False

        ()

ANSIBLE_COW_WHITELIST で利用するキャラクタを指定できるのはこんな動きになっていたんですね。
実行パスは set_cowsay_info() でこんな感じでセットされます。

set_cowsay_info
# 探索するcowsay実行パスのリスト
b_COW_PATHS = (
    b"/usr/bin/cowsay",
    b"/usr/games/cowsay",
    b"/usr/local/bin/cowsay",  # BSD path for cowsay
    b"/opt/local/bin/cowsay",  # MacPorts path for cowsay
)

()

    def set_cowsay_info(self):
        # ANSIBLE_NOCOWSで無効化されていなければ実行パスをセットする
        if not C.ANSIBLE_NOCOWS:

            # b_COW_PATHSに含まれるもののうち、存在するものをb_cowsayにセット
            for b_cow_path in b_COW_PATHS:
                if os.path.exists(b_cow_path):
                    self.b_cowsay = b_cow_path

こんな感じで、 b_cowsay に実行パスが設定されるとめでたくcowsay状態となります。

また、ここまでに base.yml から以下3つのコンフィグ変数が使われています。

  • ANSIBLE_COW_SELECTION
  • ANSIBLE_COW_WHITELIST
  • ANSIBLE_NOCOWS
base.ymlより抜粋
ANSIBLE_COW_SELECTION:
  name: Cowsay filter selection
  default: default
  description: This allows you to chose a specific cowsay stencil for the banners or use 'random' to cycle through them.
  env: [{name: ANSIBLE_COW_SELECTION}]
  ini:
  - {key: cow_selection, section: defaults}
ANSIBLE_COW_WHITELIST:
  name: Cowsay filter whitelist
  default: ['bud-frogs', 'bunny', 'cheese', 'daemon', 'default', 'dragon', 'elephant-in-snake', 'elephant', 'eyes', 'hellokitty', 'kitty', 'luke-koala', 'meow', 'milk', 'moofasa', 'moose', 'ren', 'sheep', 'small', 'stegosaurus', 'stimpy', 'supermilker', 'three-eyes', 'turkey', 'turtle', 'tux', 'udder', 'vader-koala', 'vader', 'www']
  description: White list of cowsay templates that are 'safe' to use, set to empty list if you want to enable all installed templates.
  env: [{name: ANSIBLE_COW_WHITELIST}]
  ini:
  - {key: cow_whitelist, section: defaults}
  type: list
  yaml: {key: display.cowsay_whitelist}
ANSIBLE_NOCOWS:
  name: Suppress cowsay output
  default: False
  description: If you have cowsay installed but want to avoid the 'cows' (why????), use this.
  env: [{name: ANSIBLE_NOCOWS}]
  ini:
  - {key: nocows, section: defaults}
  type: boolean
  yaml: {key: display.i_am_no_fun}

実行

読み込んだcowsayパスを元に、 banner でのバナー出力を行います。

banner
    def banner(self, msg, color=None, cows=True):
        '''
        Prints a header-looking line with cowsay or stars wit hlength depending on terminal width (3 minimum)
        '''
        # cowsayの実行パスが指定されており、明示的に実行しないよう(cows=False)に
        # されていなければbanner_cowsayを利用する
        if self.b_cowsay and cows:
            try:
                # cowsayでバナー出力したらバナー処理が終了する
                self.banner_cowsay(msg)
                return
            except OSError:
                # 何かあったらメッセージを出して通常バナーで処理を続行
                self.warning("somebody cleverly deleted cowsay or something during the PB run.  heh.")

        # ここから通常バナー処理
        msg = msg.strip()
        star_len = self.columns - len(msg)
        if star_len <= 3:
            star_len = 3
        stars = u"*" * star_len
        self.display(u"\n%s %s" % (msg, stars), color=color)

なるほどー、と思ったのですが、通常のバナー処理よりもcowsayによるバナー処理を割り込ませているんですね。
割り込んだ処理がうまく行ったらそこで return にすることで、通常バナーを抑止しつつ、何か失敗してもその先で通常バナーが出るようになっているようです。

では最後に肝心の banner_cowsay を見てみましょう。

banner_cowsay
    def banner_cowsay(self, msg, color=None):
        # メッセージ中にある "[" "]" のセットを削除する
        if u": [" in msg:
            msg = msg.replace(u"[", u"")
            if msg.endswith(u"]"):
                msg = msg[:-1]

        # 実行するcowsayコマンドを組み立て
        runcmd = [self.b_cowsay, b"-W", b"60"]

        # cowsayキャラクタを選択
        # randomの場合はcows_availableからランダムに選ぶ
        if self.noncow:
            thecow = self.noncow
            if thecow == 'random':
                thecow = random.choice(list(self.cows_available))
            runcmd.append(b'-f')
            runcmd.append(to_bytes(thecow))

        # メッセージ本文をコマンドに追加して実行
        runcmd.append(to_bytes(msg))
        cmd = subprocess.Popen(runcmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
        (out, err) = cmd.communicate()

        # displayメソッドでcowsayコマンドから受け取った出力を画面へ出力
        self.display(u"%s\n" % to_text(out), color=color)

まぁ、ここ自体はそこまで面白いものではないです。
最初の "[" "]" を削除するくだりはぱっと見で意味不明でしたが、冷静に読むと特に何ということはない処理ですね。

というような感じでcowsayバージョンの出力が得られるようになっていました。

読み込み(echo-sd版)

読み込み部分はこんな感じに落ち着きました。
echo-sdにはcowsayのように出力のスタイルが豊富ではないので、 cows_available に相当する部分は素直に b_ECHO_SD_STYLES にしました。

init
# echo-sd実行パスのリスト
b_ECHO_SD_PATHS = (
b"/usr/local/bin/echo-sd", # path for echo-sd
)

# echo-sdの出力スタイル一覧
b_ECHO_SD_STYLES = (
b"vertical",
b"tanzaku",
b"default",
)

class Display:

    def __init__(self, verbosity=0):

        ()

        # echo-sd実行パス
        self.b_echo_sd = None

        # echo-sdの出力スタイルをコンフィグから読み込む
        self.nonsd = C.ANSIBLE_SD_SELECTION

        # echo-sdの実行パスを決定
        self.set_echo_sd_info()

        ()

    def set_echo_sd_info(self):
        if not C.ANSIBLE_NOSD:
            for b_echo_sd_path in b_ECHO_SD_PATHS:
                if os.path.exists(b_echo_sd_path):
                    self.b_echo_sd = b_echo_sd_path

()

コードを見たらわかりますが、コンフィグの変数は

  • ANSIBLE_SD_SELECTION
    • echo-sdの出力スタイルを指定
    • randomで「通常モード(default)」「縦書きモード(vertical)」「短冊モード(tanzaku)」から選ばれる
  • ANSIBLE_NOSD
    • Trueにするとecho-sdによるバナー出力が抑制される
    • デフォルトではFalseなのでサーバ上にecho-sdがインストールされていればecho-sdスタイルで動く

の2つが使われます。

これでecho-sdの実行準備が整いました。

実行(echo-sd版)

こちらはこんな感じに。処理スタイルはそのままに、呼び出すメソッド名を変えただけですね。

banner
    def banner(self, msg, color=None, sd=True):
        '''
        Prints a header-looking line with echo-sd or stars wit hlength depending on terminal width (3 minimum)
        '''
        if self.b_echo_sd and sd:
            try:
                self.banner_echo_sd(msg)
                return
            except OSError:
                self.warning("somebody cleverly deleted echo-sd or something during the PB run.  heh.")

        ()

具体的な処理はこうなりました。基本的な流れは元のままですが、地味に調整を入れています。

banner_echo_sd
    def banner_echo_sd(self, msg, color=None):
        # メッセージ中にある "[" "]" を削除する
        if u": [" in msg:
            msg = msg.replace(u"[", u"")
            if msg.endswith(u"]"):
                msg = msg[:-1]

        # 実行するecho-sdコマンドを組み立て
        runcmd = [self.b_echo_sd]

        # echo-sdの出力スタイルを選択
        if self.nonsd:
            thesd = self.nonsd
            if thesd == 'random':
                thesd = random.choice(list(b_ECHO_SD_STYLES))

            # default以外だったらオプションを追加
            if thesd != 'default':
                runcmd.append(to_bytes("--" + thesd))

                # default以外のスタイルではメッセージに
                # "[" "]" が入っているとなぜか出力がおかしくなるので削除
                msg = msg.replace(u"[", u"")
                msg = msg.replace(u"]", u"")

        # メッセージ本文をコマンドに追加して実行
        runcmd.append(to_bytes(msg))
        cmd = subprocess.Popen(runcmd, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
        (out, err) = cmd.communicate()

        # displayメソッドでcowsayコマンドから受け取った出力を画面へ出力
        # 表示がきつくなるので1行分改行してはじめる
        print()
        self.display(u"%s\n" % to_text(out), color=color)

よくわからなかったのは「メッセージ本文に [ ] のセットが含まれると なぜか出力がおかしくなる 」点です。
この削除処理を入れない状態では、

_人人_
> PLAY [localhost]        <
> LAY [localhost]       <
> AY [localhost]       <
> Y [localhost]      <
> [localhost]      <
> ? <
> l <
> o <
> c <
> a <
> l <
> h <
> o <
> s <
> t <
> ? <
 ̄Y^Y^ ̄

というような感じになってしまい、はてな?という感じでした。
最初はPython側でメッセージ本文のパースでおかしなことになったのかと思いきや、実際に直接echo-sdで試してみると

# コマンドで直接やってもおかしい
$ echo-sd --vertical "TASK [Display message]"
_人人_
> TASK [Display message]           <
> ASK [Display message]          <
> SK [Display message]          <
> K [Display message]         <
> [Display message]         <
> ? <
> D <
> i <
> s <
> p <
> l <
> a <
> y <
>  <
> m <
> e <
> s <
> s <
> a <
> g <
> e <
> ? <
 ̄Y^Y^ ̄

# "[" "]" を除去するとちゃんと出る
$ echo-sd --vertical "TASK Display message"
_人人_
> T <
> A <
> S <
> K <
>  <
> D <
> i <
> s <
> p <
> l <
> a <
> y <
>  <
> m <
> e <
> s <
> s <
> a <
> g <
> e <
 ̄Y^Y^ ̄

ということがあり、デフォルト以外で出力する場合にはメッセージ本文から [ ] を削除しています。
echo-sdでは縦書きで出力する際にフォーマット揃えのためアルファベットを全角に変換しているようなのですが、 [ ] が入っていると正規表現かなにかがおかしくなるんでしょうか?
たぶんソース見たらわかると思うんですがとりあえず問題は回避できているので深追いはしていません。

実行サンプル

というわけでこんなものが出来上がりました。 縦書きが入るとなおウザくてたまらないですね。

$ export ANSIBLE_SD_SELECTION=random
$ ansible-playbook -i hosts echo-sd.yml

_人人人人人人人人人人_
> PLAY [localhost] <
 ̄Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^ ̄


_人人人人人人人人人人人人人_
> TASK [Gathering Facts] <
 ̄Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^Y^ ̄

ok: [localhost]

┏-┷-┓
┃ T ┃
┃ A ┃
┃ S ┃
┃ K ┃
┃  ┃
┃ D ┃
┃ i ┃
┃ s ┃
┃ p ┃
┃ l ┃
┃ a ┃
┃ y ┃
┃  ┃
┃ m ┃
┃ e ┃
┃ s ┃
┃ s ┃
┃ a ┃
┃ g ┃
┃ e ┃
┗━━┛

ok: [localhost] => {
    "msg": "突然の死"
}

_人人_
> P <
> L <
> A <
> Y <
>  <
> R <
> E <
> C <
> A <
> P <
 ̄Y^Y^ ̄

localhost                  : ok=2    changed=0    unreachable=0    failed=0

めでたしめでたし。

16
4
2

Register as a new user and use Qiita more conveniently

  1. You get articles that match your needs
  2. You can efficiently read back useful information
  3. You can use dark theme
What you can do with signing up
16
4

Delete article

Deleted articles cannot be recovered.

Draft of this article would be also deleted.

Are you sure you want to delete this article?