Collect metrics from promscale

[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 d870444ca41db0ba5be42d26786ed6ce8b2e1f80..40d49640aab47dbf3ede461976f638d15fb79f42 100644 (file)
--- a/cookbooks/prometheus/templates/default/alert_rules.yml.erb
+++ b/cookbooks/prometheus/templates/default/alert_rules.yml.erb
@@ -8,6 +8,29 @@ groups:
         for: 5m
         labels:
           alertgroup: "prometheus"
         for: 5m
         labels:
           alertgroup: "prometheus"

+  - name: apache
+    rules:
+      - alert: apache down
+        expr: apache_up == 0
+        for: 5m
+        labels:
+          alertgroup: "{{ $labels.instance }}"
+      - alert: apache workers busy
+        expr: sum(apache_workers{state="busy"}) by (instance) / sum(apache_scoreboard) by (instance) > 0.8
+        for: 5m
+        labels:
+          alertgroup: "{{ $labels.instance }}"
+        annotations:
+          busy_workers: "{{ $value | humanizePercentage }}"
+  - name: database
+    rules:
+      - alert: postgres replication delay
+        expr: pg_replication_lag_seconds > 5
+        for: 5m
+        labels:
+          alertgroup: database
+        annotations:
+          delay: "{{ $value | humanizeDuration }}"

   - name: hwmon
     rules:
       - alert: hwmon fan alarm
   - name: hwmon
     rules:
       - alert: hwmon fan alarm


@@ -104,3 +127,70 @@ groups:
           alertgroup: "{{ $labels.instance }}"
         annotations:
           new_oom_kills: "{{ $value }}"
           alertgroup: "{{ $labels.instance }}"
         annotations:
           new_oom_kills: "{{ $value }}"

+  - name: network
+    rules:
+      - alert: interface transmit rate
+        expr: rate(node_network_transmit_bytes_total[1m]) / node_network_speed_bytes > 0.98
+        for: 5m
+        labels:
+          alertgroup: "{{ $labels.instance }}"
+        annotations:
+          bandwidth_used: "{{ $value | humanizePercentage }}"
+      - alert: interface receive rate
+        expr: rate(node_network_receive_bytes_total[1m]) / node_network_speed_bytes > 0.98
+        for: 5m
+        labels:
+          alertgroup: "{{ $labels.instance }}"
+        annotations:
+          bandwidth_used: "{{ $value | humanizePercentage }}"
+      - alert: interface transmit errors
+        expr: rate(node_network_transmit_errs_total[1m]) / rate(node_network_transmit_packets_total[1m]) > 0.01
+        for: 5m
+        labels:
+          alertgroup: "{{ $labels.instance }}"
+        annotations:
+          error_rate: "{{ $value | humanizePercentage }}"
+      - alert: interface receive errors
+        expr: rate(node_network_receive_errs_total[1m]) / rate(node_network_receive_packets_total[1m]) > 0.01
+        for: 5m
+        labels:
+          alertgroup: "{{ $labels.instance }}"
+        annotations:
+          error_rate: "{{ $value | humanizePercentage }}"
+      - alert: conntrack entries
+        expr: node_nf_conntrack_entries / node_nf_conntrack_entries_limit > 0.8
+        for: 5m
+        labels:
+          alertgroup: "{{ $labels.instance }}"
+        annotations:
+          entries_used: "{{ $value | humanizePercentage }}"
+  - name: tile
+    rules:
+      - alert: renderd replication delay
+        expr: renderd_replication_delay > 120
+        for: 5m
+        labels:
+          alertgroup: tile
+        annotations:
+          delay: "{{ $value | humanizeDuration }}"
+      - alert: missed tile rate
+        expr: sum(rate(modtile_http_response_total{code="404"}[5m])) by (instance) / sum(rate(modtile_http_response_total[5m])) by (instance) > 0.05
+        for: 5m
+        labels:
+          alertgroup: tile
+        annotations:
+          miss_rate: "{{ $value | humanizePercentage }}"
+  - name: time
+    rules:
+      - alert: clock not synchronising
+        expr: min_over_time(node_timex_sync_status[1m]) == 0 and node_timex_maxerror_seconds >= 16
+        for: 5m
+        labels:
+          alertgroup: "{{ $labels.instance }}"
+      - alert: clock skew detected
+        expr: (node_timex_offset_seconds > 0.05 and deriv(node_timex_offset_seconds[5m]) >= 0) or (node_timex_offset_seconds < -0.05 and deriv(node_timex_offset_seconds[5m]) <= 0)
+        for: 5m
+        labels:
+          alertgroup: "{{ $labels.instance }}"
+        annotations:
+          skew: "{{ with printf \"node_timex_offset_seconds{instance='%s'}\" $labels.instance | query }} {{ . | humanizeDuration }}{{ end }}"