This repository has been archived on 2024-02-11. You can view files and clone it, but cannot push or open issues or pull requests.
valinor/kubernetes/apps/monitoring/kube-prometheus-stack/app/prometheusrules/minio.yaml

32 lines
1.1 KiB
YAML
Raw Normal View History

---
# yaml-language-server: $schema=https://kubernetes-schemas.devbu.io/monitoring.coreos.com/prometheusrule_v1.json
apiVersion: monitoring.coreos.com/v1
kind: PrometheusRule
metadata:
2023-10-31 14:57:47 -05:00
name: minio-rules
namespace: monitoring
labels:
prometheus: k8s
role: alert-rules
spec:
groups:
- name: minio-alerts
rules:
- alert: NodesOffline
2023-10-31 15:04:11 -05:00
expr: avg_over_time(minio_cluster_nodes_offline_total{job="scrapeconfig/monitoring/minio-erebor"}[5m]) > 0
for: 10m
labels:
severity: warn
annotations:
summary: "Node down in MinIO deployment"
description: "Node(s) in cluster {{ $labels.instance }} offline for more than 5 minutes"
- alert: DisksOffline
2023-10-31 15:08:29 -05:00
expr: avg_over_time(minio_cluster_drive_offline_total{job="scrapeconfig/monitoring/minio-erebor"}[5m]) > 0
for: 10m
labels:
severity: warn
annotations:
summary: "Disks down in MinIO deployment"
description: "Disks(s) in cluster {{ $labels.instance }} offline for more than 5 minutes"