Sienna Meridian Satterwhite
3fc54c8851
28 alert rules across 9 PrometheusRule files covering infrastructure (Longhorn, cert-manager), data (PostgreSQL, OpenBao, OpenSearch), storage (SeaweedFS), devtools (Gitea), identity (Hydra, Kratos), media (LiveKit), and mesh (Linkerd golden signals for all services). Severity routing: critical alerts fire to Matrix + email, warnings to Matrix only (AlertManager config updated in separate commit).
47 lines
1.6 KiB
YAML
47 lines
1.6 KiB
YAML
apiVersion: monitoring.coreos.com/v1
|
|
kind: PrometheusRule
|
|
metadata:
|
|
name: seaweedfs-alerts
|
|
namespace: storage
|
|
labels:
|
|
role: alert-rules
|
|
spec:
|
|
groups:
|
|
- name: seaweedfs
|
|
rules:
|
|
- alert: SeaweedFSMasterDown
|
|
expr: up{job=~".*seaweedfs.*", app="seaweedfs-master"} == 0
|
|
for: 2m
|
|
labels:
|
|
severity: critical
|
|
annotations:
|
|
summary: "SeaweedFS master is down"
|
|
description: "SeaweedFS master {{ $labels.namespace }}/{{ $labels.pod }} is down."
|
|
|
|
- alert: SeaweedFSVolumeDown
|
|
expr: up{job=~".*seaweedfs.*", app="seaweedfs-volume"} == 0
|
|
for: 2m
|
|
labels:
|
|
severity: critical
|
|
annotations:
|
|
summary: "SeaweedFS volume server is down"
|
|
description: "SeaweedFS volume server {{ $labels.namespace }}/{{ $labels.pod }} is down."
|
|
|
|
- alert: SeaweedFSFilerDown
|
|
expr: up{job=~".*seaweedfs.*", app="seaweedfs-filer"} == 0
|
|
for: 2m
|
|
labels:
|
|
severity: critical
|
|
annotations:
|
|
summary: "SeaweedFS filer is down"
|
|
description: "SeaweedFS filer {{ $labels.namespace }}/{{ $labels.pod }} is down."
|
|
|
|
- alert: SeaweedFSDiskLow
|
|
expr: (seaweedfs_disk_free_bytes / (seaweedfs_disk_free_bytes + seaweedfs_disk_used_bytes)) < 0.15
|
|
for: 5m
|
|
labels:
|
|
severity: warning
|
|
annotations:
|
|
summary: "SeaweedFS disk space low"
|
|
description: "SeaweedFS disk on {{ $labels.instance }} has less than 15% free space."
|