diff --git a/dist/rules/kubernetes/kubestate-exporter.yml b/dist/rules/kubernetes/kubestate-exporter.yml index 9014275..3f9dc6a 100644 --- a/dist/rules/kubernetes/kubestate-exporter.yml +++ b/dist/rules/kubernetes/kubestate-exporter.yml @@ -266,7 +266,7 @@ groups: description: "Kubernetes Job {{ $labels.namespace }}/{{ $labels.job_name }} did not complete in time.\n VALUE = {{ $value }}\n LABELS = {{ $labels }}" - alert: KubernetesApiServerErrors - expr: 'sum(rate(apiserver_request_total{job="apiserver",code=~"^(?:5..)$"}[1m])) / sum(rate(apiserver_request_total{job="apiserver"}[1m])) * 100 > 3' + expr: 'sum(rate(apiserver_request_total{job="apiserver",code=~"(?:5..)"}[1m])) by (instance, job) / sum(rate(apiserver_request_total{job="apiserver"}[1m])) by (instance, job) * 100 > 3' for: 2m labels: severity: critical @@ -302,7 +302,7 @@ groups: description: "A client certificate used to authenticate to the apiserver is expiring in less than 24.0 hours.\n VALUE = {{ $value }}\n LABELS = {{ $labels }}" - alert: KubernetesApiServerLatency - expr: 'histogram_quantile(0.99, sum(rate(apiserver_request_duration_seconds_bucket{subresource!="log",verb!~"^(?:CONNECT|WATCHLIST|WATCH|PROXY)$"} [10m])) WITHOUT (instance, resource)) > 1' + expr: 'histogram_quantile(0.99, sum(rate(apiserver_request_duration_seconds_bucket{verb!~"(?:CONNECT|WATCHLIST|WATCH|PROXY)"} [10m])) WITHOUT (subresource)) > 1' for: 2m labels: severity: warning