Make Elasticsearch restarts always rolling

fixes #343
NETWAYS · Oct 24, 2024 · b57009e · b57009e
1 parent 65dfa4f
commit b57009e
Show file tree

Hide file tree

Showing 5 changed files with 178 additions and 131 deletions.
diff --git a/roles/elasticsearch/handlers/main.yml b/roles/elasticsearch/handlers/main.yml
@@ -1,11 +1,10 @@
 ---
 # handlers file for elasticsearch
 - name: Restart Elasticsearch
-  ansible.builtin.service:
-    name: elasticsearch
-    state: restarted
-    daemon_reload: yes
+  ansible.builtin.include_tasks: restart_elasticsearch.yml
+  with_items: "{{ groups[elasticstack_elasticsearch_group_name] }}"
   when:
+    - "hostvars[item].inventory_hostname == inventory_hostname"
     - elasticsearch_enable | bool
     - not elasticsearch_freshstart.changed | bool
     - not elasticsearch_freshstart_security.changed | bool

diff --git a/roles/elasticsearch/handlers/restart_elasticsearch.yml b/roles/elasticsearch/handlers/restart_elasticsearch.yml
@@ -0,0 +1,12 @@
+---
+
+- name: Check for running Elasticsearch service
+  ansible.builtin.systemd:
+    name: elasticsearch
+  register: elasticsearch_running
+
+- name: Include rolling stop
+  ansible.builtin.include_tasks: tasks/elasticsearch-rolling-stop.yml
+
+- name: Include rolling start
+  ansible.builtin.include_tasks: tasks/elasticsearch-rolling-start.yml
diff --git a/roles/elasticsearch/tasks/elasticsearch-rolling-start.yml b/roles/elasticsearch/tasks/elasticsearch-rolling-start.yml
@@ -0,0 +1,83 @@
+# Ansible
+#
+# Rolling Upgrade of Elasticsearch with security on
+# Source from: author: Jeff Steinmetz, @jeffsteinmetz; Bin Li, @holysoros
+# Modifications: author: Daniel Neuberger @netways.de
+# More modifications: NETWAYS Professional Services GmbH
+# latest tested with Ansible 2.9 and later
+
+---
+
+# For now we support upgrade only for clusters with security enabled
+# If you positively need support for safely upgrading clusters without security,
+# feel free to open an issue at https://github.com/NETWAYS/ansible-collection-elasticstack/issues
+
+- name: Start elasticsearch
+  ansible.builtin.service:
+    name: elasticsearch
+    enabled: yes
+    state: started
+  when:
+    - elasticsearch_running.status.ActiveState == "active"
+    - not elasticsearch_unsafe_upgrade_restart | bool
+
+- name: Restart elasticsearch (fast, for non-prod)
+  ansible.builtin.service:
+    name: elasticsearch
+    enabled: yes
+    state: restarted
+  when:
+    - elasticsearch_running.status.ActiveState == "active"
+    - elasticsearch_unsafe_upgrade_restart | bool
+
+- name: Wait for elasticsearch node to come back up if it was stopped
+  ansible.builtin.wait_for:
+    host: "{{ elasticsearch_api_host }}"
+    port: "{{ elasticstack_elasticsearch_http_port }}"
+    delay: 30
+
+- name: Confirm the node joins the cluster # noqa: risky-shell-pipe
+  ansible.builtin.shell: >
+    if test -n "$(ps -p $$ | grep bash)"; then set -o pipefail; fi;
+    curl
+    -k
+    -u elastic:{{ elasticstack_password.stdout }}
+    -s
+    -m 2
+    '{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cat/nodes?h=name'
+    | grep
+    -E
+    '^{{ elasticsearch_nodename }}$'
+  register: result
+  until: result.rc == 0
+  retries: 200
+  delay: 3
+  changed_when: false
+
+- name: Enable shard allocation for the cluster
+  ansible.builtin.uri:
+    url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cluster/settings"
+    method: PUT
+    body: '{ "persistent": { "cluster.routing.allocation.enable": null }}'
+    body_format: json
+    user: elastic
+    password: "{{ elasticstack_password.stdout }}"
+    validate_certs: no
+  register: response
+  # next line is boolean not string, so no quotes around true
+  # use python truthiness
+  until: "response.json.acknowledged == true"
+  retries: 5
+  delay: 30
+
+- name: Wait for cluster health to return to yellow or green
+  ansible.builtin.uri:
+    url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cluster/health"
+    method: GET
+    user: elastic
+    password: "{{ elasticstack_password.stdout }}"
+    validate_certs: no
+  register: response
+  until: "response.json.status == 'yellow' or response.json.status == 'green'"
+  retries: 5
+  delay: 30
diff --git a/roles/elasticsearch/tasks/elasticsearch-rolling-stop.yml b/roles/elasticsearch/tasks/elasticsearch-rolling-stop.yml
@@ -0,0 +1,76 @@
+# Ansible
+#
+# Rolling Upgrade of Elasticsearch with security on
+# Source from: author: Jeff Steinmetz, @jeffsteinmetz; Bin Li, @holysoros
+# Modifications: author: Daniel Neuberger @netways.de
+# More modifications: NETWAYS Professional Services GmbH
+# latest tested with Ansible 2.9 and later
+
+---
+
+# For now we support upgrade only for clusters with security enabled
+# If you positively need support for safely upgrading clusters without security,
+# feel free to open an issue at https://github.com/NETWAYS/ansible-collection-elasticstack/issues
+- name: Set connection protocol to https
+  ansible.builtin.set_fact:
+    elasticsearch_http_protocol: "https"
+
+# Usually we should not need this step. It's only there to recover from broken upgrade plays
+# Without this step the cluster would never recover and the play would always fail
+- name: Enable shard allocation for the cluster
+  ansible.builtin.uri:
+    url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cluster/settings"
+    method: PUT
+    body: '{ "persistent": { "cluster.routing.allocation.enable": null }}'
+    body_format: json
+    user: elastic
+    password: "{{ elasticstack_password.stdout }}"
+    validate_certs: no
+  register: response
+  # next line is boolean not string, so no quotes around true
+  # use python truthiness
+  until: "response.json.acknowledged == true"
+  retries: 5
+  delay: 30
+
+  # this step is key!!!  Don't restart more nodes
+  # until all shards have completed recovery
+- name: Wait for cluster health to return to green
+  ansible.builtin.uri:
+    url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cluster/health"
+    method: GET
+    user: elastic
+    password: "{{ elasticstack_password.stdout }}"
+    validate_certs: no
+  register: response
+  until: "response.json.status == 'green'"
+  retries: 50
+  delay: 30
+
+# Disabling shard allocation right after enabling it seems redundant. Please see above for details.
+- name: Disable shard allocation for the cluster
+  ansible.builtin.uri:
+    url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cluster/settings"
+    method: PUT
+    body: '{ "persistent": { "cluster.routing.allocation.enable": "none" }}'
+    body_format: json
+    user: elastic
+    password: "{{ elasticstack_password.stdout }}"
+    validate_certs: no
+
+- name: Stop non essential indexing to speed up shard recovery
+  ansible.builtin.uri:
+    url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_flush"
+    method: POST
+    user: elastic
+    password: "{{ elasticstack_password.stdout }}"
+    validate_certs: no
+  failed_when: false
+
+- name: Shutdown elasticsearch service
+  ansible.builtin.service:
+    name: elasticsearch
+    enabled: yes
+    state: stopped
+  when:
+    - not elasticsearch_unsafe_upgrade_restart | bool
diff --git a/roles/elasticsearch/tasks/elasticsearch-rolling-upgrade.yml b/roles/elasticsearch/tasks/elasticsearch-rolling-upgrade.yml
@@ -71,65 +71,8 @@
     - groups[elasticstack_elasticsearch_group_name] | length > 1
   block:
 
-    # Usually we should not need this step. It's only there to recover from broken upgrade plays
-    # Without this step the cluster would never recover and the play would always fail
-    - name: Enable shard allocation for the cluster
-      ansible.builtin.uri:
-        url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cluster/settings"
-        method: PUT
-        body: '{ "persistent": { "cluster.routing.allocation.enable": null }}'
-        body_format: json
-        user: elastic
-        password: "{{ elasticstack_password.stdout }}"
-        validate_certs: no
-      register: response
-      # next line is boolean not string, so no quotes around true
-      # use python truthiness
-      until: "response.json.acknowledged == true"
-      retries: 5
-      delay: 30
-
-      # this step is key!!!  Don't restart more nodes
-      # until all shards have completed recovery
-    - name: Wait for cluster health to return to green
-      ansible.builtin.uri:
-        url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cluster/health"
-        method: GET
-        user: elastic
-        password: "{{ elasticstack_password.stdout }}"
-        validate_certs: no
-      register: response
-      until: "response.json.status == 'green'"
-      retries: 50
-      delay: 30
-
-    # Disabling shard allocation right after enabling it seems redundant. Please see above for details.
-    - name: Disable shard allocation for the cluster
-      ansible.builtin.uri:
-        url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cluster/settings"
-        method: PUT
-        body: '{ "persistent": { "cluster.routing.allocation.enable": "none" }}'
-        body_format: json
-        user: elastic
-        password: "{{ elasticstack_password.stdout }}"
-        validate_certs: no
-
-    - name: Stop non essential indexing to speed up shard recovery
-      ansible.builtin.uri:
-        url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_flush"
-        method: POST
-        user: elastic
-        password: "{{ elasticstack_password.stdout }}"
-        validate_certs: no
-      failed_when: false
-
-    - name: Shutdown elasticsearch service
-      ansible.builtin.service:
-        name: elasticsearch
-        enabled: yes
-        state: stopped
-      when:
-        - not elasticsearch_unsafe_upgrade_restart | bool
+    - name: Include rolling stop
+      ansible.builtin.include_tasks: elasticsearch-rolling-stop.yml
 
     - name: Update Elasticsearch - rpm with managed repositories
       ansible.builtin.package:
@@ -147,72 +90,6 @@
         - ansible_os_family == "Debian" or
           not elasticstack_full_stack | bool
 
-    - name: Start elasticsearch
-      ansible.builtin.service:
-        name: elasticsearch
-        enabled: yes
-        state: started
-      when:
-        - elasticsearch_running.status.ActiveState == "active"
-        - not elasticsearch_unsafe_upgrade_restart | bool
-
-    - name: Restart elasticsearch (fast, for non-prod)
-      ansible.builtin.service:
-        name: elasticsearch
-        enabled: yes
-        state: restarted
-      when:
-        - elasticsearch_running.status.ActiveState == "active"
-        - elasticsearch_unsafe_upgrade_restart | bool
-
-    - name: Wait for elasticsearch node to come back up if it was stopped
-      ansible.builtin.wait_for:
-        host: "{{ elasticsearch_api_host }}"
-        port: "{{ elasticstack_elasticsearch_http_port }}"
-        delay: 30
-
-    - name: Confirm the node joins the cluster # noqa: risky-shell-pipe
-      ansible.builtin.shell: >
-        if test -n "$(ps -p $$ | grep bash)"; then set -o pipefail; fi;
-        curl
-        -k
-        -u elastic:{{ elasticstack_password.stdout }}
-        -s
-        -m 2
-        '{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cat/nodes?h=name'
-        | grep
-        -E
-        '^{{ elasticsearch_nodename }}$'
-      register: result
-      until: result.rc == 0
-      retries: 200
-      delay: 3
-      changed_when: false
-
-    - name: Enable shard allocation for the cluster
-      ansible.builtin.uri:
-        url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cluster/settings"
-        method: PUT
-        body: '{ "persistent": { "cluster.routing.allocation.enable": null }}'
-        body_format: json
-        user: elastic
-        password: "{{ elasticstack_password.stdout }}"
-        validate_certs: no
-      register: response
-      # next line is boolean not string, so no quotes around true
-      # use python truthiness
-      until: "response.json.acknowledged == true"
-      retries: 5
-      delay: 30
+    - name: Include rolling start
+      ansible.builtin.include_tasks: elasticsearch-rolling-start.yml
 
-    - name: Wait for cluster health to return to yellow or green
-      ansible.builtin.uri:
-        url: "{{ elasticsearch_http_protocol }}://{{ elasticsearch_api_host }}:{{ elasticstack_elasticsearch_http_port }}/_cluster/health"
-        method: GET
-        user: elastic
-        password: "{{ elasticstack_password.stdout }}"
-        validate_certs: no
-      register: response
-      until: "response.json.status == 'yellow' or response.json.status == 'green'"
-      retries: 5
-      delay: 30