From: Tom Hughes <tom@compton.nu>
Date: Tue, 21 Mar 2023 17:34:01 +0000 (+0000)
Subject: Reduce sensitivity of CPU pressure alerts
X-Git-Url: https://git.openstreetmap.org./chef.git/commitdiff_plain/3eb308cdf3235e6d34cba81d1c88e3ab6ceaac9f?ds=sidebyside

Reduce sensitivity of CPU pressure alerts
---

diff --git a/cookbooks/prometheus/templates/default/alert_rules.yml.erb b/cookbooks/prometheus/templates/default/alert_rules.yml.erb
index cb8d91e99..c34647330 100644
--- a/cookbooks/prometheus/templates/default/alert_rules.yml.erb
+++ b/cookbooks/prometheus/templates/default/alert_rules.yml.erb
@@ -90,8 +90,8 @@ groups:
   - name: cpu
     rules:
       - alert: cpu pressure
-        expr: rate(node_pressure_cpu_waiting_seconds_total[5m]) > 0.6
-        for: 15m
+        expr: rate(node_pressure_cpu_waiting_seconds_total[5m]) > 0.75
+        for: 60m
         labels:
           alertgroup: "{{ $labels.instance }}"
         annotations: