Tweak MachineCapacityLow alert, lower repeat interval on Slack to 1h.

control-plane/roles/monitoring/templates/prometheus-stack-values.yaml

@@ -121,6 +121,7 @@ alertmanager:
     - name: 'null'
 {% if monitoring_slack_api_url is defined and monitoring_slack_notification_channel is defined %}
     - name: slack
+      repeat_interval: 1h
       slack_configs:
       - channel: "{{ monitoring_slack_notification_channel }}"
         send_resolved: true
@@ -207,7 +208,7 @@ additionalPrometheusRulesMap:
         annotations:
           description: "Partition {{ $labels.partition }} has {{ $value }} DEAD Machines"
       - alert: MachineCapacityLow
-        expr: (avg(metal_partition_capacity_total{size!="unknown"})by (partition, size) - avg(metal_partition_capacity_free{size!="unknown"}) by (partition, size) ) / avg(metal_partition_capacity_total{size!="unknown"})by (partition, size)  *100 > 80
+        expr: avg(metal_partition_capacity_free{size!="unknown"} > 10) by (partition, size) / avg(metal_partition_capacity_total{size!="unknown"} > 10) by (partition, size) < 0.10
         for: 10m
         labels:
           severity: "warning"