Added alerting.restarts.enabled

nano-byte · Jan 17, 2024 · d7aa745 · d7aa745
1 parent 9d8a9b7
commit d7aa745
Show file tree

Hide file tree

Showing 4 changed files with 15 additions and 0 deletions.
diff --git a/charts/generic-service/README.md b/charts/generic-service/README.md
@@ -141,6 +141,7 @@ app:
 | `alerting.enabled`                              | `false`                     | Deploys Prometheus alert rule for issues like like unavailable pods or high memory use                   |
 | `alerting.pod.maxStartupSeconds`                | `120`                       | The maximum amount of time a Pod is allowed to take for startup                                          |
 | `alerting.pod.maxAgeSeconds`                    |                             | The maximum allowed age of a `Pod` in seconds (useful to ensure regular deployments)                     |
+| `alerting.restarts.enabled`                     | `true`                      | Deploys Prometheus alert rule for unexpected container restarts                                          |
 | `alerting.memory.enabled`                       | `true`                      | Enables alerts relating to memory usage                                                                  |
 | `alerting.memory.maxUsageFactor`                | `0.9`                       | The maximum usage factor of the memory limit (between `0` and `1`)                                       |
 | `alerting.memory.quotaBufferFactor`             | `1.0`                       | Multiplied with `resources.*.memory` to determine minimum allowed unused memory quota in namespace       |

diff --git a/charts/generic-service/templates/alerts.yaml b/charts/generic-service/templates/alerts.yaml
@@ -45,6 +45,7 @@ spec:
         {{- end }}
       {{- end }}
 
+      {{- if .Values.alerting.restarts.enabled }}
         - alert: ContainerRestart
           # Avoid constantly retriggering during crash loops by comparing over interval slightly longer than CrashLoopBackOff upper limit (5m).
           # Don't trigger during startup grace period (service might just be waiting for dependencies).
@@ -56,6 +57,7 @@ spec:
             topic: availability
           annotations: {{- include "generic-service.alert-annotations" . | nindent 12 }} crash/restart
             description: '{{"{{ $labels.pod }}"}} has crashed/restarted.'
+      {{- end }}
 
       {{- if .Values.alerting.pod.maxAgeSeconds }}
         - alert: PodTooOld

diff --git a/charts/generic-service/values.schema.json b/charts/generic-service/values.schema.json
@@ -784,6 +784,16 @@
           },
           "additionalProperties": false
         },
+        "restarts": {
+          "type": "object",
+          "properties": {
+            "enabled": {
+              "type": "boolean",
+              "default": true,
+              "description": "Deploys Prometheus alert rule for unexpected container restarts"
+            }
+          }
+        },
         "memory": {
           "type": "object",
           "properties": {

diff --git a/charts/generic-service/values.yaml b/charts/generic-service/values.yaml
@@ -154,6 +154,8 @@ alerting:
   pod:
     maxStartupSeconds: 120
     maxAgeSeconds:
+  restarts:
+    enabled: true
   memory:
     enabled: true
     maxUsageFactor: 0.9