From 01e7ab043cba1145096fa0f6475ad6875f3c2421 Mon Sep 17 00:00:00 2001
From: Lionel H <lionel.hubaut@tessares.net>
Date: Mon, 25 Mar 2019 15:50:17 +0100
Subject: [PATCH] reboot: not correctly checking host reachability.

Come across the issue where after that the reboot has completed host was
not yet reachable. This is because the host was still booting hence
refusing SSH connection even if SSH service was up and running. wait_for
is the old way to check host reachability and it produced unreliable
results, because the system needs to be booted and we need to make sure
we can execute command over it before continuing the next task.

wait_for_connection does that pretty automatically, it wait for the host
to be reachable and ensure we can run command on it.

Signed-off-by: Lionel H <lionel.hubaut@tessares.net>
---
 README.md         |  1 +
 defaults/main.yml |  5 ++++-
 tasks/main.yml    | 25 +++----------------------
 3 files changed, 8 insertions(+), 23 deletions(-)

diff --git a/README.md b/README.md
index 536ee31..cdc45a9 100644
--- a/README.md
+++ b/README.md
@@ -18,6 +18,7 @@ A role for rebooting hosts.
 |---------------------------|-----------------------------------------------------------|--------------------------------------------|
 | `reboot_message`          | Reboot message for the logs                               | 'Reboot by Ansible'                        |
 | `reboot_wait`             | Wait for hosts to come back online?                       | 'yes'                                      |
+| `reboot_connect_timeout`  | Connection timeout before retrying.                       | 20                                         |
 | `reboot_wait_host`        | Host to check                                             | `ansible_ssh_host` or `inventory_hostname` |
 | `reboot_wait_port`        | Port to check                                             | `ansible_ssh_port` or 22                   |
 | `reboot_wait_regex`       | String to match in the socket connection. (ex. `OpenSSH`) | /                                          |
diff --git a/defaults/main.yml b/defaults/main.yml
index 411d5b8..19cb93f 100644
--- a/defaults/main.yml
+++ b/defaults/main.yml
@@ -7,6 +7,9 @@ reboot_message: 'Reboot by Ansible'
 # Wait for host?
 reboot_wait: 'yes'
 
+# Maximum number of seconds to wait for a connection to happen before closing and retrying.
+reboot_connect_timeout: 20
+
 # Time to wait before polling the host
 reboot_wait_delay: 10
 
@@ -20,4 +23,4 @@ reboot_wait_ctimeout: 5
 reboot_interval: 'no'
 
 # Interval before next task
-reboot_interval_seconds: 0
+reboot_interval_seconds: 0
\ No newline at end of file
diff --git a/tasks/main.yml b/tasks/main.yml
index 1708142..5d44cc9 100644
--- a/tasks/main.yml
+++ b/tasks/main.yml
@@ -7,31 +7,12 @@
   poll: 0
   ignore_errors: yes
 
-  # Make sure ansible_ssh_host has the correct value inside the local_action module
-- name: Set default reboot_wait_host
-  become: no
-  set_fact:
-    reboot_wait_host: "{{ ansible_ssh_host | default(inventory_hostname) }}"
-  when: (reboot_wait|bool) and reboot_wait_host is undefined
-
-  # Make sure ansible_ssh_port has the correct value inside the local_action module
-- name: Set default reboot_wait_port
-  become: no
-  set_fact:
-    reboot_wait_port: "{{ ansible_ssh_port | default(22) }}"
-  when: (reboot_wait|bool) and reboot_wait_port is undefined
-
 - name: Wait for host
-  become: no
-  wait_for:
-    host: "{{ reboot_wait_host }}"
-    port: "{{ reboot_wait_port }}"
-    state: started
+  wait_for_connection:
+    connect_timeout: "{{ reboot_connect_timeout }}"
+    sleep: "{{ reboot_wait_ctimeout }}"
     delay: "{{ reboot_wait_delay }}"
     timeout: "{{ reboot_wait_timeout }}"
-    search_regex: "{{ reboot_wait_regex | default(omit) }}"
-    connect_timeout: "{{ reboot_wait_ctimeout }}"
-  delegate_to: localhost
   when: reboot_wait|bool
 
 - name: Interval before next task