]> git.openstreetmap.org Git - chef.git/commitdiff
Reduce sensitivity of web error rate alert on idle servers master
authorTom Hughes <tom@compton.nu>
Tue, 4 Mar 2025 12:54:55 +0000 (12:54 +0000)
committerTom Hughes <tom@compton.nu>
Tue, 4 Mar 2025 12:54:55 +0000 (12:54 +0000)
cookbooks/prometheus/templates/default/alert_rules.yml.erb

index 305afbd903a44e3745ce3d3eb5dc62d51895a2b7..27cb72e476faa8bbde0d3e8d68790c8b3d71cd7f 100644 (file)
@@ -828,7 +828,7 @@ groups:
   - name: web
     rules:
       - alert: web error rate
-        expr: sum(rate(api_call_count_total{status=~"50[0-8]|5[1-9][0-9]"}[5m])) by (instance) / sum(rate(api_call_count_total[5m])) by (instance) > 0.002 and sum(rate(api_call_count_total{status=~"50[0-8]|5[1-9][0-9]"}[5m])) by (instance) > 0.01
+        expr: sum(rate(api_call_count_total{status=~"50[0-8]|5[1-9][0-9]"}[5m])) by (instance) / sum(rate(api_call_count_total[5m])) by (instance) > 0.002 and sum(rate(api_call_count_total{status=~"50[0-8]|5[1-9][0-9]"}[5m])) by (instance) > 0.05
         for: 5m
         labels:
           alertgroup: web