Add an alert for the mail queue

[chef.git] / cookbooks / prometheus / templates / default / alert_rules.yml.erb

diff --git a/cookbooks/prometheus/templates/default/alert_rules.yml.erb b/cookbooks/prometheus/templates/default/alert_rules.yml.erb
index 7d7fa4abfdeed6c429523e163f33570b5e083abd..9a3609e03a1876f10a99f4bcfac63033e100b645 100644 (file)
--- a/cookbooks/prometheus/templates/default/alert_rules.yml.erb
+++ b/cookbooks/prometheus/templates/default/alert_rules.yml.erb
@@ -25,12 +25,12 @@ groups:
   - name: cpu
     rules:
       - alert: cpu pressure
-        expr: rate(node_pressure_memory_waiting_seconds_total[5m]) > 0.3
+        expr: rate(node_pressure_cpu_waiting_seconds_total[5m]) > 0.3
         for: 15m
         labels:
           alertgroup: "{{ $labels.instance }}"
         annotations:
-          major_page_faults: "{{ $value | humanizePercentage }}"
+          pressure: "{{ $value | humanizePercentage }}"
   - name: database
     rules:
       - alert: postgres replication delay
@@ -95,13 +95,13 @@ groups:
           in_max_volts: "{{ with printf \"node_hwmon_in_max_volts{instance='%s',chip='%s',sensor='%s'}\" $labels.instance $labels.chip $labels.sensor | query }}{{ . | first | value | humanize }}V{{end}}"
   - name: io
     rules:
-      - alert: cpu pressure
+      - alert: io pressure
         expr: rate(node_pressure_io_waiting_seconds_total[5m]) > 0.6
         for: 60m
         labels:
           alertgroup: "{{ $labels.instance }}"
         annotations:
-          major_page_faults: "{{ $value | humanizePercentage }}"
+          pressure: "{{ $value | humanizePercentage }}"
   - name: ipmi
     rules:
       - alert: ipmi fan alarm
@@ -130,6 +130,15 @@ groups:
         for: 5m
         labels:
           alertgroup: "{{ $labels.instance }}"
+  - name: mail
+    rules:
+      - alert: mail queue length
+        expr: exim_queue > exim_queue_limit
+        for: 60m
+        labels:
+          alertgroup: mail
+        annotations:
+          queue_length: "{{ $value }}"
   - name: mdadm
     rules:
       - alert: mdadm array inactive
@@ -177,7 +186,7 @@ groups:
         labels:
           alertgroup: "{{ $labels.instance }}"
         annotations:
-          major_page_faults: "{{ $value | humanizePercentage }}"
+          pressure: "{{ $value | humanizePercentage }}"
       - alert: oom kill detected
         expr: increase(node_vmstat_oom_kill[1m]) > 0
         for: 0m