diff options
Diffstat (limited to 'vnfs/DAaaS/prometheus-operator/templates/alertmanager/rules/kubernetes-system.yaml')
| -rw-r--r-- | vnfs/DAaaS/prometheus-operator/templates/alertmanager/rules/kubernetes-system.yaml | 117 |
1 files changed, 117 insertions, 0 deletions
diff --git a/vnfs/DAaaS/prometheus-operator/templates/alertmanager/rules/kubernetes-system.yaml b/vnfs/DAaaS/prometheus-operator/templates/alertmanager/rules/kubernetes-system.yaml new file mode 100644 index 00000000..78f90b79 --- /dev/null +++ b/vnfs/DAaaS/prometheus-operator/templates/alertmanager/rules/kubernetes-system.yaml @@ -0,0 +1,117 @@ +# Generated from 'kubernetes-system' group from https://raw.githubusercontent.com/coreos/prometheus-operator/master/contrib/kube-prometheus/manifests/prometheus-rules.yaml +{{- if and .Values.defaultRules.create }} +apiVersion: {{ printf "%s/v1" (.Values.prometheusOperator.crdApiGroup | default "monitoring.coreos.com") }} +kind: PrometheusRule +metadata: + name: {{ printf "%s-%s" (include "prometheus-operator.fullname" .) "kubernetes-system" | trunc 63 | trimSuffix "-" }} + labels: + app: {{ template "prometheus-operator.name" . }} +{{ include "prometheus-operator.labels" . | indent 4 }} +{{- if .Values.defaultRules.labels }} +{{ toYaml .Values.defaultRules.labels | indent 4 }} +{{- end }} +{{- if .Values.defaultRules.annotations }} + annotations: +{{ toYaml .Values.defaultRules.annotations | indent 4 }} +{{- end }} +spec: + groups: + - name: kubernetes-system + rules: + - alert: KubeNodeNotReady + annotations: + message: '{{`{{ $labels.node }}`}} has been unready for more than an hour.' + runbook_url: https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubenodenotready + expr: kube_node_status_condition{job="kube-state-metrics",condition="Ready",status="true"} == 0 + for: 1h + labels: + severity: warning + - alert: KubeVersionMismatch + annotations: + message: There are {{`{{ $value }}`}} different versions of Kubernetes components running. + runbook_url: https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubeversionmismatch + expr: count(count(kubernetes_build_info{job!="kube-dns"}) by (gitVersion)) > 1 + for: 1h + labels: + severity: warning + - alert: KubeClientErrors + annotations: + message: Kubernetes API server client '{{`{{ $labels.job }}`}}/{{`{{ $labels.instance }}`}}' is experiencing {{`{{ printf "%0.0f" $value }}`}}% errors.' + runbook_url: https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubeclienterrors + expr: |- + (sum(rate(rest_client_requests_total{code=~"5.."}[5m])) by (instance, job) + / + sum(rate(rest_client_requests_total[5m])) by (instance, job)) + * 100 > 1 + for: 15m + labels: + severity: warning + - alert: KubeClientErrors + annotations: + message: Kubernetes API server client '{{`{{ $labels.job }}`}}/{{`{{ $labels.instance }}`}}' is experiencing {{`{{ printf "%0.0f" $value }}`}} errors / second. + runbook_url: https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubeclienterrors + expr: sum(rate(ksm_scrape_error_total{job="kube-state-metrics"}[5m])) by (instance, job) > 0.1 + for: 15m + labels: + severity: warning + - alert: KubeletTooManyPods + annotations: + message: Kubelet {{`{{ $labels.instance }}`}} is running {{`{{ $value }}`}} Pods, close to the limit of 110. + runbook_url: https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubelettoomanypods + expr: kubelet_running_pod_count{job="kubelet"} > 110 * 0.9 + for: 15m + labels: + severity: warning + - alert: KubeAPILatencyHigh + annotations: + message: The API server has a 99th percentile latency of {{`{{ $value }}`}} seconds for {{`{{ $labels.verb }}`}} {{`{{ $labels.resource }}`}}. + runbook_url: https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubeapilatencyhigh + expr: cluster_quantile:apiserver_request_latencies:histogram_quantile{job="apiserver",quantile="0.99",subresource!="log",verb!~"^(?:LIST|WATCH|WATCHLIST|PROXY|CONNECT)$"} > 1 + for: 10m + labels: + severity: warning + - alert: KubeAPILatencyHigh + annotations: + message: The API server has a 99th percentile latency of {{`{{ $value }}`}} seconds for {{`{{ $labels.verb }}`}} {{`{{ $labels.resource }}`}}. + runbook_url: https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubeapilatencyhigh + expr: cluster_quantile:apiserver_request_latencies:histogram_quantile{job="apiserver",quantile="0.99",subresource!="log",verb!~"^(?:LIST|WATCH|WATCHLIST|PROXY|CONNECT)$"} > 4 + for: 10m + labels: + severity: critical + - alert: KubeAPIErrorsHigh + annotations: + message: API server is returning errors for {{`{{ $value }}`}}% of requests. + runbook_url: https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubeapierrorshigh + expr: |- + sum(rate(apiserver_request_count{job="apiserver",code=~"^(?:5..)$"}[5m])) without(instance, pod) + / + sum(rate(apiserver_request_count{job="apiserver"}[5m])) without(instance, pod) * 100 > 10 + for: 10m + labels: + severity: critical + - alert: KubeAPIErrorsHigh + annotations: + message: API server is returning errors for {{`{{ $value }}`}}% of requests. + runbook_url: https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubeapierrorshigh + expr: |- + sum(rate(apiserver_request_count{job="apiserver",code=~"^(?:5..)$"}[5m])) without(instance, pod) + / + sum(rate(apiserver_request_count{job="apiserver"}[5m])) without(instance, pod) * 100 > 5 + for: 10m + labels: + severity: warning + - alert: KubeClientCertificateExpiration + annotations: + message: Kubernetes API certificate is expiring in less than 7 days. + runbook_url: https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubeclientcertificateexpiration + expr: histogram_quantile(0.01, sum by (job, le) (rate(apiserver_client_certificate_expiration_seconds_bucket{job="apiserver"}[5m]))) < 604800 + labels: + severity: warning + - alert: KubeClientCertificateExpiration + annotations: + message: Kubernetes API certificate is expiring in less than 24 hours. + runbook_url: https://github.com/kubernetes-monitoring/kubernetes-mixin/tree/master/runbook.md#alert-name-kubeclientcertificateexpiration + expr: histogram_quantile(0.01, sum by (job, le) (rate(apiserver_client_certificate_expiration_seconds_bucket{job="apiserver"}[5m]))) < 86400 + labels: + severity: critical +{{- end }}
\ No newline at end of file |