Ansible：playbookで指定したサーバ再起動を正常処理したい

Question

ansibleで作成したWindowsサーバー(aws:ec2)のサーバ名(hostname)を変更し、サーバの再起動を実行したい内容のplaybookを作成しました。

・ansible 2.6.1
・Ansibleインスタンス：Centos7.5
・管理対象インスタンス：Win2012R2

# cat hosts
localhost ansible_host=127.0.0.1 ansible_connection=local

[windows]
ec2-11-222-33-444.ap-northeast-1.compute.amazonaws.com ansible_ssh_host=11.222.33.444 ansible_ssh_user=Administrator ansible_ssh_pass=xxxxxxxxxx ansible_ssh_port=5985 ansible_connection=winrm

$ cat site.yml
---
- hosts: windows
  roles:
    - ini_set

.
初期設定を実施しているplaybook ↓ です。ホスト名変更以外にも複数の処理を実行していますがスルーで。

$ cat roles/ini_set/tasks/main.yml
  - name: Install IIS
    win_feature:
      name: "Web-Server"
      state: present
      restart: yes
      include_sub_features: yes
      include_management_tools: yes

  - name: Setup Timezones
    win_timezone:
      timezone: "Tokyo Standard Time"

  - name: disabled windows firewall service
    win_service:
      name: "Windows Firewall"
      start_mode: disabled

  - name: Change the hostname to new_hostname
    win_hostname:
      name: win-test-svr
    register: win_hostname

  - name: Reboot
    win_reboot:
    when: win_hostname.reboot_required

.
playbookを実行した時の状況です。

# ansible-playbook -l windows site.yml -vvvvv

（途中省略）

TASK [ini_set : Change the hostname to new_hostname] *******************************************************************
task path: /etc/ansible/roles/ini_set/tasks/main.yml:18
Using module file /usr/lib/python2.7/site-packages/ansible/modules/windows/win_hostname.ps1
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
checking if winrm_host 11.222.33.444 is an IPv6 address
<11.222.33.444> WINRM CONNECT: transport=plaintext endpoint=http://11.222.33.444:5985/wsman
<11.222.33.444> WINRM OPEN SHELL: B627CE20-ADDD-480A-B4F3-2E11D28CB49C
EXEC (via pipeline wrapper)
<11.222.33.444> WINRM EXEC 'PowerShell' ['-NoProfile', '-NonInteractive', '-ExecutionPolicy', 'Unrestricted', '-']
<11.222.33.444> WINRM RESULT u'<Response code 0, out "{"changed":true,"old", err "">'
<11.222.33.444> WINRM CLOSE SHELL: B627CE20-ADDD-480A-B4F3-2E11D28CB49C
changed: [ec2-11-222-33-444.ap-northeast-1.compute.amazonaws.com] => {
    "changed": true,
    "old_name": "WIN-G9P8SPSN37M",
    "reboot_required": true
}

TASK [ini_set : Reboot] ************************************************************************************************
task path: /etc/ansible/roles/ini_set/tasks/main.yml:23
skipping: [ec2-11-222-33-444.ap-northeast-1.compute.amazonaws.com] => {
    "changed": false,
    "skip_reason": "Conditional result was False"
}
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
checking if winrm_host 11.222.33.444 is an IPv6 address
<11.222.33.444> WINRM CONNECT: transport=plaintext endpoint=http://11.222.33.444:5985/wsman
<11.222.33.444> WINRM OPEN SHELL: EC7761B5-D68A-4676-A865-78B579B5E212
EXEC (via pipeline wrapper)
<11.222.33.444> WINRM EXEC 'PowerShell' ['-NoProfile', '-NonInteractive', '-ExecutionPolicy', 'Unrestricted', '-EncodedCommand', u'KABHAGUAdAAtAFcAbQBpAE8AYgBqAGUAYwB0ACAALQBDAGwAYQBzAHMATgBhAG0AZQAgAFcAaQBuADMAMgBfAE8AcABlAHIAYQB0AGkAbgBnAFMAeQBzAHQAZQBtACkALgBMAGEAcwB0AEIAbwBvAHQAVQBwAFQAaQBtAGUA']
<11.222.33.444> WINRM RESULT u'<Response code 0, out "20180724105429.49088", err "">'
rebooting server
EXEC (via pipeline wrapper)
<11.222.33.444> WINRM EXEC 'PowerShell' ['-NoProfile', '-NonInteractive', '-ExecutionPolicy', 'Unrestricted', '-EncodedCommand', u'cwBoAHUAdABkAG8AdwBuACAALwByACAALwB0ACAAMgAgAC8AYwAgACIAUgBlAGIAbwBvAHQAIABpAG4AaQB0AGkAYQB0AGUAZAAgAGIAeQAgAEEAbgBzAGkAYgBsAGUALgAiAA==']
<11.222.33.444> WINRM RESULT u'<Response code 0, out "", err "">'
attempting to get system uptime
attempting to get system uptime
attempting to get system uptime
　　　・
　　　・　　　　※「attempting to get system uptime」が続く
　　　・
attempting to get system uptime
attempting to get system uptime
attempting to get system uptime
<inventory_hostname> WINRM CLOSE SHELL: EC7761B5-D68A-4676-A865-78B579B5E212
fatal: [ec2-11-222-33-444.ap-northeast-1.compute.amazonaws.com]: FAILED! => {
    "changed": false,
    "elapsed": 600,
    "msg": "timed out waiting for reboot uptime check success: Invalid settings supplied for _extras: Requested option _extras was not defined in configuration",
    "rebooted": true
}
META: ran handlers

PLAY RECAP *************************************************************************************************************
ec2-11-222-33-444.ap-northeast-1.compute.amazonaws.com : ok=5    changed=4    unreachable=0    failed=1

この後、対象のサーバーにリモートデスクトップログインしてみると、hostname は意図したものに変更されていました。
処理としてはFatalになっているので、上記の「attempting to get system uptime」を回避したいのですが、その方法がわからず…。

足りない情報は追記していきますので、よろしくお願いします。

Answer 1

このGithubページに記載していましたが、ansible 2.6.1 の不具合だった様です。
こちらのページを参考に、2.6.2 にアップグレードしたら解決しました。

$ sudo pip uninstall ansible
$ sudo pip install ansible\==2.6.2
$ ansible --version
ansible 2.6.2

# ansible-playbook -l windows site.yml -vvvvv

（途中省略）

TASK [ini_set : Reboot] ************************************************************************************************
task path: /etc/ansible/roles/ini_set/tasks/main.yml:37
 [WARNING]: Since Ansible 2.5, shutdown_timeout_sec is no longer used with win_reboot

<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
EXEC (via pipeline wrapper)
rebooting server
EXEC (via pipeline wrapper)
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
EXEC (via pipeline wrapper)
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
EXEC (via pipeline wrapper)
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
EXEC (via pipeline wrapper)
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
attempting to get system uptime
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
EXEC (via pipeline wrapper)
<11.222.33.444> ESTABLISH WINRM CONNECTION FOR USER: Administrator on PORT 5985 TO 11.222.33.444
attempting post-reboot test command 'whoami'
EXEC (via pipeline wrapper)
changed: [ec2-11-222-33-444.ap-northeast-1.compute.amazonaws.com] => {
    "changed": true,
    "elapsed": 43,
    "rebooted": true
}
META: ran handlers
META: ran handlers

PLAY RECAP *************************************************************************************************************
ec2-11-222-33-444.ap-northeast-1.compute.amazonaws.com : ok=9    changed=4    unreachable=0    failed=0