Improve alerting for degraded RAID arrays

- adjust rule for inactive arrays to reflect it not being about degraded arrays - add rule for arrays with less active disks than expected, suggesting a degraded state samber/awesome-prometheus-alerts#395 Signed-off-by: Georg Pfuetzenreuter <[email protected]>
openSUSE · Aug 7, 2024 · 3f50b91 · 3f50b91
1 parent ed3ae01
commit 3f50b91
Showing 1 changed file with 20 additions and 0 deletions.
diff --git a/salt/files/prometheus/alerts/raid.yml b/salt/files/prometheus/alerts/raid.yml
@@ -17,6 +17,26 @@ groups:
           group_left (nodename)
           node_uname_info{nodename=~".+"}
         for: 0m
+        labels:
+          severity: critical
+        annotations:
+          summary: >-
+            Inactive RAID array on {{ $labels.instance }}
+          description: |
+            RAID array {{ $labels.device }} on {{ $labels.instance }} is in an inactive state.
+            VALUE = {{ $value }}
+            LABELS = {{ $labels }}
+
+      - alert: RAID Array Degraded
+        expr: >-
+          (
+            (
+              node_md_disks_required - on(device, instance)
+              node_md_disks{state="active"}
+            )
+            > 0
+          )
+        for: 0m
         labels:
           severity: critical
         annotations: