From 40cea7a5f0b47a9e89d2651194c4936721d9919f Mon Sep 17 00:00:00 2001
From: Ryan Kraus <rmkraus@gmail.com>
Date: Tue, 29 Jun 2021 00:46:08 -0400
Subject: [PATCH] QA fixes to destroy cluster command.

This is for issue #72
---
 app/cli/destroy                              |   1 +
 app/playbooks/destroy.d/cluster/destroy.yaml | 313 -------------------
 app/playbooks/destroy.d/cluster/destroy.yml  | 131 ++++++++
 app/roles/rhcos-images/templates/wipe.ign.j2 |   3 +-
 devel.env                                    |   4 +-
 5 files changed, 136 insertions(+), 316 deletions(-)
 create mode 120000 app/cli/destroy
 delete mode 100755 app/playbooks/destroy.d/cluster/destroy.yaml
 create mode 100755 app/playbooks/destroy.d/cluster/destroy.yml

diff --git a/app/cli/destroy b/app/cli/destroy
new file mode 120000
index 0000000..1fa21ed
--- /dev/null
+++ b/app/cli/destroy
@@ -0,0 +1 @@
+.command
\ No newline at end of file
diff --git a/app/playbooks/destroy.d/cluster/destroy.yaml b/app/playbooks/destroy.d/cluster/destroy.yaml
deleted file mode 100755
index 08ca963..0000000
--- a/app/playbooks/destroy.d/cluster/destroy.yaml
+++ /dev/null
@@ -1,313 +0,0 @@
-#!/usr/bin/env ansible-playbook
-- name: Are you sure?
-  hosts: all
-  any_errors_fatal: yes
-  max_fail_percentage: 0
-  serial: 100%
-  gather_facts: no
-
-  vars_prompt:
-    - name: confirm
-      prompt: This will destroy the running cluster and you will lose all data. Are you sure? [yes to continue]
-      private: no
-
-  tasks:
-    - name: Validate user input
-      fail:
-        msg: "Action cancelled."
-      when: 'confirm != "yes"'
-
-- import_playbook: /app/playbooks/util_vm_facts.yml
-
-- name: Destroy cluster nodes
-  hosts: cluster
-  become: no
-  gather_facts: no
-
-  pre_tasks:
-    - include_role:
-        name: management
-        defaults_from: main.yml
-        tasks_from: "poweroff/{{ mgmt_provider }}.yml"
-        vars_from: "{{ mgmt_provider }}.yml"
-        handlers_from: "{{ mgmt_provider }}.yml"
-      vars:
-        management_hostname: "{{ mgmt_hostname }}"
-        management_user: "{{ mgmt_user }}"
-        management_pass: "{{ mgmt_password }}"
-
-  roles:
-    - name: pxelinux-kickstarts
-      pxelinux_kickstarts_pxe_server: "{{ groups.bastion_hosts.0 }}"
-      pxelinux_kickstarts_content_server: "http://{{ hostvars[groups.bastion_hosts.0].ansible_host }}:8081"
-      pxelinux_kickstarts_host_role: "{{ node_role }}"
-      pxelinux_kickstarts_install_disk: "{{ install_disk }}"
-      pxelinux_kickstarts_profile: wipe
-
-  tasks:
-    - include_role:
-        name: management
-        defaults_from: main.yml
-        tasks_from: "netboot/{{ mgmt_provider }}.yml"
-        vars_from: "{{ mgmt_provider }}.yml"
-        handlers_from: "{{ mgmt_provider }}.yml"
-      vars:
-        management_hostname: "{{ mgmt_hostname }}"
-        management_user: "{{ mgmt_user }}"
-        management_pass: "{{ mgmt_password }}"
-
-    - name: wait for nodes to start wiping
-      shell: ping -c 1 {{ ansible_host }}
-      delegate_to: "{{ groups.bastion_hosts[0] }}"
-      become: no
-      register: node_ping
-      until: "node_ping is not failed"
-      retries: 30
-      delay: 10
-      changed_when: no
-
-- name: Terminate the Bootstrap node
-  hosts: bootstrap
-  become: no
-  gather_facts: no
-
-  tasks:
-    - include_role:
-        name: management
-        defaults_from: main.yml
-        tasks_from: "poweroff/{{ mgmt_provider }}.yml"
-        vars_from: "{{ mgmt_provider }}.yml"
-        handlers_from: "{{ mgmt_provider }}.yml"
-      vars:
-        management_hostname: "{{ mgmt_hostname }}"
-        management_user: "{{ mgmt_user }}"
-        management_pass: "{{ mgmt_password }}"
-
-- name: Clean coreos install repos
-  hosts: bootstrap
-  become: yes
-  gather_facts: no
-
-  tasks:
-    - name: erase stored images and igniton configs
-      shell: rm -f /var/www/html/*
-
-    - name: erase pxelinux boot kickstarts
-      file:
-        path: /var/lib/tftpboot/pxelinux.cfg
-        state: absent
-
-    - name: erase uefi grub kickstarts
-      file:
-        path: /var/lib/tftpboot/uefi
-        state: absent
-
-    - name: erase coreos boot images
-      file:
-        path: /var/lib/tftpboot/rhcos
-        state: absent
-
-- name: Remove cluster dhcp records
-  hosts: cluster, management
-  gather_facts: no
-  serial: 1
-
-  roles:
-    - name: dhcp
-      dhcp_present: no
-
-- name: Remove ignored MAC Addresses
-  hosts: localhost
-  gather_facts: no
-  serial: 1
-
-  vars:
-    dns_domain: "{{ cluster_name }}.{{ cluster_domain }}"
-    dns_reverse_domain: "{{ reverse_ptr_zone }}"
-
-  tasks:
-    - name: configure extra records
-      include_role:
-        name: dhcp
-      vars:
-        dhcp_name: "{{ node.name }}"
-        dhcp_mac_address: "{{ node.mac }}"
-        dhcp_ignore: yes
-        dhcp_present: no
-      loop: "{{ ignored_macs }}"
-      loop_control:
-        loop_var: node
-
-- name: Remove cluster node records
-  hosts: cluster, management
-  gather_facts: no
-  serial: 1
-
-  vars:
-    dns_domain: "{{ cluster_name }}.{{ cluster_domain }}"
-    dns_reverse_domain: "{{ reverse_ptr_zone }}"
-
-  roles:
-    - name: dns
-      dns_present: no
-
-- name: Remove Load Balanced DNS entries
-  hosts: loadbalancer
-  gather_facts: no
-  serial: 1
-
-  vars:
-    dns_domain: "{{ cluster_name }}.{{ cluster_domain }}"
-    dns_reverse_domain: "{{ reverse_ptr_zone }}"
-
-  roles:
-    - name: dns
-      dns_present: no
-      dns_hostname: "api"
-      dns_reverse: no
-    - name: dns
-      dns_present: no
-      dns_hostname: "api-int"
-      dns_reverse: no
-    - name: dns
-      dns_present: no
-      dns_hostname: "*.apps"
-      dns_reverse: no
-    - name: dns
-      dns_present: no
-      dns_hostname: "loadbalancer"
-
-- name: Remove etcd DNS entries
-  hosts: control_plane
-  gather_facts: no
-  serial: 1
-
-  vars:
-    dns_domain: "{{ cluster_name }}.{{ cluster_domain }}"
-    dns_reverse_domain: "{{ reverse_ptr_zone }}"
-
-  roles:
-    - name: dns
-      dns_present: no
-      dns_hostname: "etcd-{{ cp_node_id }}"
-      dns_reverse: no
-    - name: dns
-      dns_present: no
-      dns_hostname: "_etcd-server-ssl.tcp"
-      dns_type: 'SRV'
-      dns_value: "0 10 2380 etcd-{{ cp_node_id }}.{{ cluster_name }}.{{ cluster_domain }}."
-
-- name: Remove extra DNS records
-  hosts: localhost
-  gather_facts: no
-  serial: 1
-
-  vars:
-    dns_domain: "{{ cluster_name }}.{{ cluster_domain }}"
-    dns_reverse_domain: "{{ reverse_ptr_zone }}"
-
-  tasks:
-    - name: configure extra records
-      include_role:
-        name: dns
-      vars:
-        dns_present: no
-        dns_hostname: "{{ node.name }}"
-        dns_value: "{{ node.ip }}"
-      loop: "{{ extra_nodes }}"
-      loop_control:
-        loop_var: node
-
-- name: Purge openshift installation resource caches
-  host: localhost
-  gather_facts: no
-
-  tasks:
-    - name: remove installer data cache
-      file:
-        path: /data/openshift-installer
-        state: absent
-
-    - name: remove ip address database
-      file:
-        path: /data/ip_addresses
-        state: absent
-
-- name: Uninstall cluster load balancer
-  host: bastion
-  gather_facts: no
-
-  tasks:
-    - name: shutdown the load balancer service
-      systemd:
-        name: haproxy
-        state: stopped
-        enabled: no
-        force: yes
-
-    - name: shutdown the VIP service
-      systemd:
-        name: keepalived
-        state: stopped
-        enabled: no
-        force: yes
-
-    - name: delete load balancer configuration
-      file:
-        path: /etc/haproxy/haproxy.cfg
-        state: absent
-
-    - name: delete the vip configuration
-      file:
-        path: /etc/keepalived/keepalived.conf
-        state: absent
-
-    - name: remove vip cached password
-      file:
-        path: /data/keepalived.pass
-        state: absent
-      delegate_to: localhost
-
-- name: Clean firewall configuration
-  host: bastion
-  gather_facts: no
-  become: yes
-
-  roles:
-    - name: router
-      router_all_interfaces: "{{ all_interfaces }}"
-      router_lan_interfaces: "{{ lan_interfaces }}"
-      router_lan_address: "{{ hostvars.lan.ansible_host }}"
-      router_lan_subnet: "{{ subnet }}/{{ subnet_mask }}"
-      router_subnet_mask: "{{ subnet_mask }}"
-      router_wan_interface: "{{ wan_interface }}"
-      router_loadbalancer: "{{ loadbalancer_vip }}"
-      router_dns_forwarders: "{{ dns_forwarders }}"
-      allowed_services:
-        - "SSH to Bastion"
-
-- name: Clean cockpit interface
-  host: bastion
-  gather_facts: no
-  become: yes
-
-  tasks:
-    - name: remove faros page from cockpit
-      file:
-        path: /usr/local/share/cockpit/faros
-        state: absent
-
-- name: Restart network services
-  hosts: bastion
-  become: yes
-
-  tasks:
-    - name: restart named
-      systemd:
-        name: named
-        state: restarted
-
-    - name: restart dhcpd
-      systemd:
-        name: dhcpd
-        state: restarted
diff --git a/app/playbooks/destroy.d/cluster/destroy.yml b/app/playbooks/destroy.d/cluster/destroy.yml
new file mode 100755
index 0000000..9355a15
--- /dev/null
+++ b/app/playbooks/destroy.d/cluster/destroy.yml
@@ -0,0 +1,131 @@
+#!/usr/bin/env ansible-playbook
+- name: Are you sure?
+  hosts: all
+  any_errors_fatal: yes
+  max_fail_percentage: 0
+  serial: 100%
+  gather_facts: no
+
+  vars_prompt:
+    - name: confirm
+      prompt: This will destroy the running cluster and you will lose all data. Are you sure? [yes to continue]
+      private: no
+
+  tasks:
+    - name: Validate user input
+      fail:
+        msg: "Action cancelled."
+      when: 'confirm != "yes"'
+
+- import_playbook: /app/playbooks/util_vm_facts.yml
+
+- name: Destroy cluster nodes
+  hosts: cluster:!bootstrap
+  become: no
+  gather_facts: no
+  any_errors_fatal: yes
+  max_fail_percentage: 0
+  serial: 1
+
+  pre_tasks:
+    - include_role:
+        name: management
+        defaults_from: main.yml
+        tasks_from: "poweroff/{{ mgmt_provider }}.yml"
+        vars_from: "{{ mgmt_provider }}.yml"
+        handlers_from: "{{ mgmt_provider }}.yml"
+      vars:
+        management_hostname: "{{ mgmt_hostname }}"
+        management_user: "{{ mgmt_user }}"
+        management_pass: "{{ mgmt_password }}"
+
+  roles:
+    - name: pxelinux-kickstarts
+      pxelinux_kickstarts_pxe_server: "{{ groups.bastion_hosts.0 }}"
+      pxelinux_kickstarts_content_server: "http://{{ hostvars[groups.bastion_hosts.0].ansible_host }}:8081"
+      pxelinux_kickstarts_host_role: "{{ node_role }}"
+      pxelinux_kickstarts_install_disk: "{{ install_disk }}"
+      pxelinux_kickstarts_profile: wipe
+
+  tasks:
+    - include_role:
+        name: management
+        defaults_from: main.yml
+        tasks_from: "netboot/{{ mgmt_provider }}.yml"
+        vars_from: "{{ mgmt_provider }}.yml"
+        handlers_from: "{{ mgmt_provider }}.yml"
+      vars:
+        management_hostname: "{{ mgmt_hostname }}"
+        management_user: "{{ mgmt_user }}"
+        management_pass: "{{ mgmt_password }}"
+
+- name: Wait for cluster nodes to start wiping
+  hosts: cluster:!bootstrap
+  become: no
+  gather_facts: no
+  any_errors_fatal: yes
+  max_fail_percentage: 0
+
+  tasks:
+    - name: wait for nodes to start wiping
+      shell: ping -c 1 {{ ansible_host }}
+      delegate_to: "{{ groups.bastion_hosts[0] }}"
+      become: no
+      register: node_ping
+      until: "node_ping is not failed"
+      retries: 30
+      delay: 10
+      changed_when: no
+
+- name: Terminate the Bootstrap node
+  hosts: bootstrap
+  become: no
+  gather_facts: no
+
+  tasks:
+    - include_role:
+        name: management
+        defaults_from: main.yml
+        tasks_from: "poweroff/{{ mgmt_provider }}.yml"
+        vars_from: "{{ mgmt_provider }}.yml"
+        handlers_from: "{{ mgmt_provider }}.yml"
+      vars:
+        management_hostname: "{{ mgmt_hostname }}"
+        management_user: "{{ mgmt_user }}"
+        management_pass: "{{ mgmt_password }}"
+
+- name: Clean cockpit interface
+  hosts: bastion
+  gather_facts: no
+  become: yes
+
+  tasks:
+    - name: remove faros page from cockpit
+      file:
+        path: /usr/local/share/cockpit/faros
+        state: absent
+
+
+- name: Wait for hosts to finish disk wipe
+  hosts: cluster:!bootstrap
+  become: no
+  gather_facts: no
+
+  tasks:
+    - name: create tmp cache dir
+      tempfile:
+        state: directory
+      register: cache
+      delegate_to: localhost
+
+    - name: wait for node to power off
+      shell: |
+        ilorest --cache-dir="{{ cache.path }}" login "{{ mgmt_hostname }}" -u '{{ mgmt_user }}' -p '{{ mgmt_password }}' &>/dev/null
+        ilorest --cache-dir="{{ cache.path }}" --nologo get --selector=ComputerSystem PowerState --json
+      register: powercheck
+      changed_when: False
+      delegate_to: localhost
+      retries: 60
+      delay: 10
+      until: "(powercheck.stdout | from_json)['PowerState'] == 'Off'"
+
diff --git a/app/roles/rhcos-images/templates/wipe.ign.j2 b/app/roles/rhcos-images/templates/wipe.ign.j2
index a2e4ff6..e22a1a0 100644
--- a/app/roles/rhcos-images/templates/wipe.ign.j2
+++ b/app/roles/rhcos-images/templates/wipe.ign.j2
@@ -1,4 +1,5 @@
-tion": {
+{
+  "ignition": {
     "version": "3.2.0"
   },
   "systemd": {
diff --git a/devel.env b/devel.env
index 1707140..bc427aa 100644
--- a/devel.env
+++ b/devel.env
@@ -1,5 +1,5 @@
-WAN_INT=ens1
-BASTION_IP_ADDR=192.168.5.16
+WAN_INT=eno2
+BASTION_IP_ADDR=192.168.8.1
 BASTION_SSH_USER=core
 BASTION_HOST_NAME=bastion
 CLUSTER_DOMAIN=faros.site