Fix scaling (#5889) (#5911)

* etcd: etcd-events doesn't depend on etcd_cluster_setup

Signed-off-by: Etienne Champetier <champetier.etienne@gmail.com>

* etcd: remove condition already present on include_tasks

Signed-off-by: Etienne Champetier <champetier.etienne@gmail.com>

* etcd: fix scaling up

Signed-off-by: Etienne Champetier <champetier.etienne@gmail.com>

* etcd: use *access_addresses, do not delegate to etcd[0]

We want to wait for the full cluster to be healthy,
so use all the cluster addresses
Also we should be able to run the playbook when etcd[0] is down
(not tested), so do not delegate to etcd[0]

Signed-off-by: Etienne Champetier <champetier.etienne@gmail.com>

* etcd: use failed_when for health check

unhealthy cluster is expected on first run, so use failed_when
instead of ignore_errors to remove scary red messages

Also use run_once

Signed-off-by: Etienne Champetier <champetier.etienne@gmail.com>

* kubernetes/preinstall: ensure ansible_fqdn is up to date after changing /etc/hosts

Signed-off-by: Etienne Champetier <champetier.etienne@gmail.com>

* kubernetes/master: regenerate apiserver cert if needed

Signed-off-by: Etienne Champetier <champetier.etienne@gmail.com>
(cherry picked from commit a35b6dc1af)

roles/etcd/tasks/configure.yml

@@ -2,9 +2,10 @@
 - name: Configure | Check if etcd cluster is healthy
   shell: "{{ bin_dir }}/etcdctl --endpoints={{ etcd_access_addresses }} cluster-health | grep -q 'cluster is healthy'"
   register: etcd_cluster_is_healthy
-  ignore_errors: true
+  failed_when: false
   changed_when: false
   check_mode: no
+  run_once: yes
   when: is_etcd_master and etcd_cluster_setup
   tags:
     - facts
@@ -16,9 +17,10 @@
 - name: Configure | Check if etcd-events cluster is healthy
   shell: "{{ bin_dir }}/etcdctl --endpoints={{ etcd_events_access_addresses }} cluster-health | grep -q 'cluster is healthy'"
   register: etcd_events_cluster_is_healthy
-  ignore_errors: true
+  failed_when: false
   changed_when: false
   check_mode: no
+  run_once: yes
   when: is_etcd_master and etcd_events_cluster_setup
   tags:
     - facts
@@ -49,22 +51,26 @@
     daemon_reload: true
   when: is_etcd_master
 
+# when scaling new etcd will fail to start
 - name: Configure | Ensure etcd is running
   service:
     name: etcd
     state: started
     enabled: yes
+  ignore_errors: "{{ etcd_cluster_is_healthy.rc == 0 }}"
   when: is_etcd_master and etcd_cluster_setup
 
+# when scaling new etcd will fail to start
 - name: Configure | Ensure etcd-events is running
   service:
     name: etcd-events
     state: started
     enabled: yes
+  ignore_errors: "{{ etcd_events_cluster_is_healthy.rc == 0 }}"
   when: is_etcd_master and etcd_events_cluster_setup
 
-- name: Configure | Check if etcd cluster is healthy
-  shell: "{{ bin_dir }}/etcdctl --no-sync --endpoints={{ etcd_client_url }} cluster-health | grep -q 'cluster is healthy'"
+- name: Configure | Wait for etcd cluster to be healthy
+  shell: "{{ bin_dir }}/etcdctl --no-sync --endpoints={{ etcd_access_addresses }} cluster-health | grep -q 'cluster is healthy'"
   register: etcd_cluster_is_healthy
   until: etcd_cluster_is_healthy.rc == 0
   retries: 4
@@ -72,7 +78,6 @@
   ignore_errors: false
   changed_when: false
   check_mode: no
-  delegate_to: "{{ groups['etcd'][0] }}"
   run_once: yes
   when:
     - is_etcd_master
@@ -84,8 +89,8 @@
     ETCDCTL_KEY_FILE: "{{ etcd_cert_dir }}/admin-{{ inventory_hostname }}-key.pem"
     ETCDCTL_CA_FILE: "{{ etcd_cert_dir }}/ca.pem"
 
-- name: Configure | Check if etcd-events cluster is healthy
-  shell: "{{ bin_dir }}/etcdctl --no-sync --endpoints={{ etcd_events_client_url }} cluster-health | grep -q 'cluster is healthy'"
+- name: Configure | Wait for etcd-events cluster to be healthy
+  shell: "{{ bin_dir }}/etcdctl --no-sync --endpoints={{ etcd_events_access_addresses }} cluster-health | grep -q 'cluster is healthy'"
   register: etcd_events_cluster_is_healthy
   until: etcd_events_cluster_is_healthy.rc == 0
   retries: 4
@@ -93,12 +98,10 @@
   ignore_errors: false
   changed_when: false
   check_mode: no
-  delegate_to: "{{ groups['etcd'][0] }}"
   run_once: yes
   when:
     - is_etcd_master
     - etcd_events_cluster_setup
-    - etcd_cluster_setup
   tags:
     - facts
   environment:
@@ -136,14 +139,10 @@
 
 - name: Configure | Join member(s) to etcd cluster one at a time
   include_tasks: join_etcd_member.yml
-  vars:
-    target_node: "{{ item }}"
   with_items: "{{ groups['etcd'] }}"
   when: inventory_hostname == item and etcd_cluster_setup and etcd_member_in_cluster.rc != 0 and etcd_cluster_is_healthy.rc == 0
 
 - name: Configure | Join member(s) to etcd-events cluster one at a time
   include_tasks: join_etcd-events_member.yml
-  vars:
-    target_node: "{{ item }}"
   with_items: "{{ groups['etcd'] }}"
   when: inventory_hostname == item and etcd_events_cluster_setup and etcd_events_member_in_cluster.rc != 0 and etcd_events_cluster_is_healthy.rc == 0

roles/etcd/tasks/join_etcd-events_member.yml

@@ -5,7 +5,6 @@
   until: member_add_result.rc == 0
   retries: 4
   delay: "{{ retry_stagger | random + 3 }}"
-  when: target_node == inventory_hostname
   environment:
     ETCDCTL_CERT_FILE: "{{ etcd_cert_dir }}/admin-{{ inventory_hostname }}.pem"
     ETCDCTL_KEY_FILE: "{{ etcd_cert_dir }}/admin-{{ inventory_hostname }}-key.pem"
@@ -21,7 +20,6 @@
           {{ etcd_member_name }}={{ etcd_events_peer_url }}
         {%- endif -%}
       {%- endfor -%}
-  when: target_node == inventory_hostname
 
 - name: Join Member | Ensure member is in etcd-events cluster
   shell: "{{ bin_dir }}/etcdctl --no-sync --endpoints={{ etcd_events_access_addresses }} member list | grep -q {{ etcd_events_access_address }}"
@@ -30,7 +28,12 @@
   check_mode: no
   tags:
     - facts
-  when: target_node == inventory_hostname
   environment:
     ETCDCTL_CERT_FILE: "{{ etcd_cert_dir }}/admin-{{ inventory_hostname }}.pem"
     ETCDCTL_KEY_FILE: "{{ etcd_cert_dir }}/admin-{{ inventory_hostname }}-key.pem"
+
+- name: Configure | Ensure etcd-events is running
+  service:
+    name: etcd-events
+    state: started
+    enabled: yes

roles/etcd/tasks/join_etcd_member.yml

@@ -5,7 +5,6 @@
   until: member_add_result.rc == 0
   retries: 4
   delay: "{{ retry_stagger | random + 3 }}"
-  when: target_node == inventory_hostname
   environment:
     ETCDCTL_CERT_FILE: "{{ etcd_cert_dir }}/admin-{{ inventory_hostname }}.pem"
     ETCDCTL_KEY_FILE: "{{ etcd_cert_dir }}/admin-{{ inventory_hostname }}-key.pem"
@@ -22,7 +21,6 @@
           {{ etcd_member_name }}={{ etcd_peer_url }}
         {%- endif -%}
       {%- endfor -%}
-  when: target_node == inventory_hostname
 
 - name: Join Member | Ensure member is in etcd cluster
   shell: "{{ bin_dir }}/etcdctl --no-sync --endpoints={{ etcd_access_addresses }} member list | grep -q {{ etcd_access_address }}"
@@ -31,8 +29,13 @@
   check_mode: no
   tags:
     - facts
-  when: target_node == inventory_hostname
   environment:
     ETCDCTL_CERT_FILE: "{{ etcd_cert_dir }}/admin-{{ inventory_hostname }}.pem"
     ETCDCTL_KEY_FILE: "{{ etcd_cert_dir }}/admin-{{ inventory_hostname }}-key.pem"
     ETCDCTL_CA_FILE: "{{ etcd_cert_dir }}/ca.pem"
+
+- name: Configure | Ensure etcd is running
+  service:
+    name: etcd
+    state: started
+    enabled: yes

roles/kubernetes/master/tasks/kubeadm-setup.yml

@@ -103,6 +103,37 @@
     - not upgrade_cluster_setup
     - kubeadm_already_run.stat.exists
 
+- name: kubeadm | Check if apiserver.crt contains all needed SANs
+  command: openssl x509 -noout -in "{{ kube_cert_dir }}/apiserver.crt" -checkip "{{ item }}"
+  with_items: "{{ apiserver_sans }}"
+  register: apiserver_sans_check
+  changed_when: "'does match certificate' not in apiserver_sans_check.stdout"
+  when:
+    - inventory_hostname == groups['kube-master']|first
+    - kubeadm_already_run.stat.exists
+
+- name: kubeadm | regenerate apiserver cert 1/2
+  file:
+    state: absent
+    path: "{{ kube_cert_dir }}/{{ item }}"
+  with_items:
+    - apiserver.crt
+    - apiserver.key
+  when:
+    - inventory_hostname == groups['kube-master']|first
+    - kubeadm_already_run.stat.exists
+    - apiserver_sans_check.changed
+
+- name: kubeadm | regenerate apiserver cert 2/2
+  command: >-
+    {{ bin_dir }}/kubeadm
+    init phase certs apiserver
+    --config={{ kube_config_dir }}/kubeadm-config.yaml
+  when:
+    - inventory_hostname == groups['kube-master']|first
+    - kubeadm_already_run.stat.exists
+    - apiserver_sans_check.changed
+
 - name: kubeadm | Initialize first master
   command: >-
     timeout -k 300s 300s

roles/kubernetes/preinstall/tasks/0090-etchosts.yml

@@ -59,3 +59,8 @@
     backup: yes
     unsafe_writes: yes
   with_dict: "{{ etc_hosts_localhosts_dict_target }}"
+
+# gather facts to update ansible_fqdn
+- name: Update facts
+  setup:
+    gather_subset: min