Merge "Update expected code for http_response"
diff --git a/cinder/meta/prometheus.yml b/cinder/meta/prometheus.yml
index aca77cc..6269b8a 100644
--- a/cinder/meta/prometheus.yml
+++ b/cinder/meta/prometheus.yml
@@ -1,6 +1,13 @@
-{% raw %}
+{%- from "cinder/map.jinja" import controller as controller with context %}
+{%- from "cinder/map.jinja" import volume as volume with context %}
+{%- set is_controller = controller.get('enabled', False) %}
+{%- set is_volume = volume.get('enabled', False) %}
+
+{%- if is_controller or is_volume %}
server:
alert:
+{%- if is_controller %}
+{%- raw %}
CinderAPIDown:
if: >-
max(openstack_api_check_status{service=~"cinder.+"}) by (service) == 0
@@ -12,4 +19,17 @@
summary: "Endpoint check for '{{ $labels.service }}' is down"
description: >-
Endpoint check for '{{ $labels.service }}' is down for 2 minutes
-{% endraw %}
\ No newline at end of file
+{%- endraw %}
+{%- endif %}
+ CinderErrorLogsTooHigh:
+ {%- set log_threshold = prometheus_server.get('alert', {}).get('CinderErrorLogsTooHigh', {}).get('var', {}).get('threshold', 0.2 ) %}
+ if: >-
+ sum(rate(log_messages{service="cinder",level=~"error|emergency|fatal"}[5m])) without (level) > {{ log_threshold }}
+{%- raw %}
+ labels:
+ severity: warning
+ service: "{{ $labels.service }}"
+ annotations:
+ summary: 'Too many errors in {{ $labels.service }} logs'
+ description: 'The rate of errors in {{ $labels.service }} logs over the last 5 minutes is too high on node {{ $labels.host }} (current value={{ $value }}, threshold={%- endraw %}{{ log_threshold }}).'
+{%- endif %}
diff --git a/metadata/service/support.yml b/metadata/service/support.yml
index d1f3754..9298460 100644
--- a/metadata/service/support.yml
+++ b/metadata/service/support.yml
@@ -14,4 +14,6 @@
grafana:
enabled: true
prometheus:
- enabled: true
\ No newline at end of file
+ enabled: true
+ telegraf:
+ enabled: true