monitoring stack images now pre-pulled

Author: wtripp180901

ansible/fatimage.yml

@@ -114,6 +114,39 @@
         slurm_exporter_state: stopped
       when: "'slurm_exporter' in group_names"
 
+- hosts: builder
+  name: Pre-pull kube-prometheus-stack images and import to k3s
+  vars:
+    image_list:
+      - { name: "docker.io/grafana/grafana", tag: "{{ grafana_image_tag }}" }
+      - { name: "quay.io/prometheus/prometheus", tag: "{{ prometheus_image_tag }}" }
+      - { name: "quay.io/prometheus/alertmanager", tag: "{{ alertmanager_image_tag }}" }
+      - { name: "quay.io/prometheus-operator/node-exporter", tag: "{{ node_exporter_image_tag }}" }
+      - { name: "quay.io/prometheus-operator/prometheus-config-reloader", tag: "{{ kube_prometheus_stack_app_version }}" }
+      - { name: "quay.io/prometheus-operator/prometheus-operator", tag: "{{ kube_prometheus_stack_app_version }}" }
+      - { name: "quay.io/kiwigrid/k8s-sidecar", tag: "{{ grafana_sidecar_image_tag }}" }
+      - { name: "registry.k8s.io/kube-state-metrics/kube-state-metrics", tag: "{{ kube_prometheus_stack_metrics_image_tag }}" }
+      - { name: "registry.k8s.io/ingress-nginx/kube-webhook-certgen", tag: "{{ kube_prometheus_stack_patch_image_tag }}" }
+  tasks:
+    - name: Pull with images with podman
+      containers.podman.podman_image:
+        name: "{{ item.name }}"
+        tag: "{{ item.tag }}"
+      loop: "{{ image_list }}"
+
+    - name: Export images to k3s
+      containers.podman.podman_save:
+        image: "{{ item.name }}:{{ item.tag }}"
+        dest: "/var/lib/rancher/k3s/agent/images/{{ item.name | regex_replace('\\/|\\.','-')}}.tar"
+      loop: "{{ image_list }}"
+
+    - name: Clean up podman images
+      containers.podman.podman_image:
+        state: absent
+        name: "{{ item.name }}"
+        tag: "{{ item.tag }}"
+      loop: "{{ image_list }}"
+
 - name: Run post.yml hook
   vars:
     appliances_environment_root: "{{ lookup('env', 'APPLIANCES_ENVIRONMENT_ROOT') }}"

ansible/roles/kube_prometheus_stack/defaults/main/helm.yml

@@ -100,6 +100,8 @@ kube_prometheus_stack_release_defaults:
     ingress:
       path: "/node/{{ groups['prometheus'].0 }}/{{ grafana_port }}"
     sidecar:
+      image:
+        tag: "{{ grafana_sidecar_image_tag }}"
       dashboards:
         searchNamespace: ALL
     grafana.ini:
@@ -147,10 +149,21 @@ kube_prometheus_stack_release_defaults:
 
 
   prometheusOperator:
+    image:
+      tag: "{{ kube_prometheus_stack_app_version }}"
+    prometheusConfigReloader:
+      image:
+        tag: "{{ kube_prometheus_stack_app_version }}"
+    admissionWebhooks:
+      patch:
+        image:
+          tag: "{{ kube_prometheus_stack_patch_image_tag }}"
     nodeSelector:
         clusterrole: "server"
 
   kube-state-metrics:
+    image:
+      tag: "{{ kube_prometheus_stack_metrics_image_tag }}"
     nodeSelector:
         clusterrole: "server"
 

ansible/roles/kube_prometheus_stack/defaults/main/main.yml

@@ -4,6 +4,7 @@
 kube_prometheus_stack_chart_repo: https://prometheus-community.github.io/helm-charts
 kube_prometheus_stack_chart_name: kube-prometheus-stack
 kube_prometheus_stack_chart_version: 59.1.0
+kube_prometheus_stack_app_version: v0.74.0
 
 # Release information
 kube_prometheus_stack_release_namespace: monitoring-system
@@ -12,6 +13,9 @@ kube_prometheus_stack_release_name: kube-prometheus-stack
 # The timeout to wait for the release to become ready
 kube_prometheus_stack_wait_timeout: 5m
 
+kube_prometheus_stack_metrics_image_tag: v2.12.0
+kube_prometheus_stack_patch_image_tag: v20221220-controller-v1.5.1-58-g787ea74b6
+
 control_ip: "{{ hostvars[groups['control'].0].ansible_host }}"
 
 grafana_auth_anonymous: false
@@ -84,6 +88,7 @@ grafana_datasources: []
 grafana_home_dashboard: 13427.json # node exporter slurm
 
 alertmanager_image_tag: v0.27.0
+grafana_sidecar_image_tag: 1.26.1
 
 node_exporter_image_tag: v1.8.2
 node_exporter_args: 

environments/common/inventory/group_vars/all/alertmanager.yml

@@ -1,3 +1,5 @@
+alertmanager_image_tag: v0.27.0
+
 alertmanager_config:
   route:
     group_by: ['...']

environments/common/inventory/group_vars/all/grafana.yml

@@ -1,4 +1,6 @@
 grafana_port: 30001
+grafana_image_tag: 11.2.2
+grafana_sidecar_image_tag: 1.26.1
 
 # Configure external address, with external URL depending on whether we are using Open Ondemand as a proxy
 grafana_api_address: "{{ hostvars[groups['prometheus'].0].internal_address }}"

environments/common/inventory/group_vars/all/monitoring.yml

@@ -4,3 +4,13 @@ kube_prometheus_stack_release_name: kube-prometheus-stack
 kube_prometheus_stack_wait_timeout: 5m
 
 # See prometheus.yml, grafana.yml and alertmanager.yml for config of individual monitoring services
+
+
+# used as tag prometheus-operator image and helpers, 
+# should be kept in sync with the appVersion provided in the Chart.yaml of whichever
+# kube_prometheus_stack_chart_version is used
+kube_prometheus_stack_app_version: v0.74.0
+
+# helper images
+kube_prometheus_stack_metrics_image_tag: v2.12.0
+kube_prometheus_stack_patch_image_tag: v20221220-controller-v1.5.1-58-g787ea74b6

environments/common/inventory/group_vars/all/node_exporter.yml

@@ -10,3 +10,5 @@ node_exporter_args:
   - --collector.cpufreq
   - --collector.diskstats
   - --collector.filesystem
+
+node_exporter_image_tag: v1.8.2

environments/common/inventory/group_vars/all/prometheus.yml

@@ -1,6 +1,8 @@
 # Must be within K3s' reserved port range (default 30000-32767)
 prometheus_port: 30000
 
+prometheus_image_tag: "v2.27.0"
+
 prometheus_db_dir: "{{ appliances_state_dir }}/prometheus"
 prometheus_storage_retention: "30d"
 prometheus_storage_retention_size: "40GB"