Docs: Clean up 'win_reboot' module docs (#46377)

* Docs: Clean up 'win_reboot' module docs

This is part of a series of module doc cleanups.

* Remove hypothetical example

lib/ansible/modules/windows/win_reboot.py

@@ -12,72 +12,96 @@ DOCUMENTATION = r'''
 module: win_reboot
 short_description: Reboot a windows machine
 description:
-     - Reboot a Windows machine, wait for it to go down, come back up, and respond to commands.
-version_added: "2.1"
+- Reboot a Windows machine, wait for it to go down, come back up, and respond to commands.
+version_added: '2.1'
 options:
   pre_reboot_delay:
     description:
-    - Seconds for shutdown to wait before requesting reboot
+    - Seconds for shutdown to wait before requesting reboot.
     type: int
     default: 2
     aliases: [ pre_reboot_delay_sec ]
   post_reboot_delay:
     description:
-    - Seconds to wait after the reboot was successful and the connection was re-established
-    - This is useful if you want wait for something to settle despite your connection already working
+    - Seconds to wait after the reboot was successful and the connection was re-established.
+    - This is useful if you want wait for something to settle despite your connection already working.
     type: int
     default: 0
     version_added: '2.4'
     aliases: [ post_reboot_delay_sec ]
   shutdown_timeout:
     description:
-    - Maximum seconds to wait for shutdown to occur
-    - Increase this timeout for very slow hardware, large update applications, etc
-    - This option has been removed since Ansible 2.5 as the win_reboot behavior has changed
+    - Maximum seconds to wait for shutdown to occur.
+    - Increase this timeout for very slow hardware, large update applications, etc.
+    - This option has been removed since Ansible 2.5 as the win_reboot behavior has changed.
     type: int
     default: 600
     aliases: [ shutdown_timeout_sec ]
   reboot_timeout:
     description:
-    - Maximum seconds to wait for machine to re-appear on the network and respond to a test command
-    - This timeout is evaluated separately for both network appearance and test command success (so maximum clock time is actually twice this value)
+    - Maximum seconds to wait for machine to re-appear on the network and respond to a test command.
+    - This timeout is evaluated separately for both network appearance and test command success (so maximum clock time is actually twice this value).
     type: int
     default: 600
     aliases: [ reboot_timeout_sec ]
   connect_timeout:
     description:
-    - Maximum seconds to wait for a single successful TCP connection to the WinRM endpoint before trying again
+    - Maximum seconds to wait for a single successful TCP connection to the WinRM endpoint before trying again.
     type: int
     default: 5
     aliases: [ connect_timeout_sec ]
   test_command:
     description:
-    - Command to expect success for to determine the machine is ready for management
+    - Command to expect success for to determine the machine is ready for management.
+    type: str
     default: whoami
   msg:
     description:
-    - Message to display to users
+    - Message to display to users.
+    type: str
     default: Reboot initiated by Ansible
 notes:
 - If a shutdown was already scheduled on the system, C(win_reboot) will abort the scheduled shutdown and enforce its own shutdown.
+- Beware that when C(win_reboot) returns, the Windows system may not have settled yet and some base services could be in limbo.
+  This can result in unexpected behavior. Check the examples for ways to mitigate this.
 - For non-Windows targets, use the M(reboot) module instead.
 author:
-    - Matt Davis (@nitzmahone)
+- Matt Davis (@nitzmahone)
 '''
 
 EXAMPLES = r'''
-# Unconditionally reboot the machine with all defaults
-- win_reboot:
+- name: Reboot the machine with all defaults
+  win_reboot:
 
-# Apply updates and reboot if necessary
-- win_updates:
-  register: update_result
-- win_reboot:
-  when: update_result.reboot_required
-
-# Reboot a slow machine that might have lots of updates to apply
-- win_reboot:
+- name: Reboot a slow machine that might have lots of updates to apply
+  win_reboot:
     reboot_timeout: 3600
+
+# Install a Windows feature and reboot if necessary
+- name: Install IIS Web-Server
+  win_feature:
+    name: Web-Server
+  register: iis_install
+
+- name: Reboot when Web-Server feature requires it
+  win_reboot:
+  when: iis_install.reboot_required
+
+# One way to ensure the system is reliable, is to add a delay before running the next task
+- name: Reboot a machine that takes time to settle after being booted
+  win_reboot:
+    post_reboot_delay: 120
+
+# Alternatively, you can set WinRM to a delayed startup
+- name: Ensure WinRM starts when the system has settled and is ready to work reliably
+  win_service:
+    name: WinRM
+    start_mode: delayed
+
+# Or you can make win_reboot validate exactly what you need to work before running the next task
+- name: Validate that the netlogon service has started, before running the next task
+  win_reboot:
+    test_command: 'exit (Get-Service -Name Netlogon).Status -ne "Running"'
 '''
 
 RETURN = r'''