[kube-monitoring] bm disk errors are warnings because of raid redundancy

sapcc · Dec 12, 2018 · 215a7fa · 215a7fa
1 parent 7eefc20
commit 215a7fa
Showing 1 changed file with 2 additions and 2 deletions.
diff --git a/system/kube-monitoring/charts/prometheus-frontend/metal-ironic.alerts b/system/kube-monitoring/charts/prometheus-frontend/metal-ironic.alerts
@@ -2,7 +2,7 @@ groups:
 - name: metal-ironic.alerts
   rules:      
   - alert: MetalIronicSensorCritical
-    expr: count(ipmi_sensor_state{type=~"(Memory|Drive Slot|Processor|Critical Interrupt)", maintenance="false", provision_state=~"(deploy|active|available)"} == 2) by (instance, type, name, manufacturer, model, provision_state, server_id, project_id)
+    expr: count(ipmi_sensor_state{type=~"(Memory|Processor|Critical Interrupt)", maintenance="false", provision_state=~"(deploy|active|available)"} == 2) by (instance, type, name, manufacturer, model, provision_state, server_id, project_id)
     for: 15m
     labels:
       severity: critical
@@ -16,7 +16,7 @@ groups:
       summary: "Hardware error for instance: {{ $labels.instance }}"
 
   - alert: MetalIronicSensorWarning
-    expr: count(ipmi_sensor_state{type=~"(Memory|Drive Slot|Processor|Critical Interrupt)", maintenance="false", provision_state=~"(manageable)"} == 2) by (instance, type, name, manufacturer, model, provision_state, server_id, project_id)
+    expr: count((ipmi_sensor_state{type=~"(Memory|Processor|Critical Interrupt)", maintenance="false", provision_state=~"(manageable)"} == 2) or (ipmi_sensor_state{type=~"(Drive Slot)", maintenance="false"} == 2)) by (instance, type, name, manufacturer, model, provision_state, server_id, project_id)
     for: 15m
     labels:
       severity: warning