add service monitoring for clickhouse (#375)

* add service monitoring for clickhouse * Update Helm documentation * add readonly user in clickhouse for monitoring * Update Helm documentation --------- Co-authored-by: CometActions <[email protected]>
comet-ml · Oct 13, 2024 · a9b5acc · a9b5acc
1 parent 6a4665d
commit a9b5acc
Show file tree

Hide file tree

Showing 4 changed files with 49 additions and 1 deletion.
diff --git a/deployment/helm_chart/opik/README.md b/deployment/helm_chart/opik/README.md
@@ -92,6 +92,10 @@ Call opik api on http://localhost:5173/api
 | clickhouse.backup.enabled | bool | `false` |  |
 | clickhouse.image | string | `"altinity/clickhouse-server:24.3.5.47.altinitystable"` |  |
 | clickhouse.logsLevel | string | `"information"` |  |
+| clickhouse.monitoring.enabled | bool | `false` |  |
+| clickhouse.monitoring.monitoringReleaseName | string | `nil` |  |
+| clickhouse.monitoring.password | string | `"opikmon"` |  |
+| clickhouse.monitoring.username | string | `"opikmon"` |  |
 | clickhouse.namePrefix | string | `"opik"` |  |
 | clickhouse.operator.enabled | bool | `true` |  |
 | clickhouse.replicasCount | int | `1` |  |

diff --git a/deployment/helm_chart/opik/templates/clickhouse-operator-install-bundle.yaml b/deployment/helm_chart/opik/templates/clickhouse-operator-install-bundle.yaml
@@ -1156,6 +1156,7 @@ spec:
 # In order to get access outside of k8s it should be exposed as:
 # kubectl --namespace prometheus port-forward service/prometheus 9090
 # and point browser to localhost:9090
+{{- if .Values.clickhouse.monitoring.enables }}
 kind: Service
 apiVersion: v1
 metadata:
@@ -1171,4 +1172,30 @@ spec:
     - port: 9999
       name: operator-metrics
   selector:
-    app: clickhouse-operator
+    app: clickhouse-operator
+{{- end }}
+---
+## ServiceMonitor
+{{- if .Values.clickhouse.monitoring.enables }}
+apiVersion: monitoring.coreos.com/v1
+kind: ServiceMonitor
+metadata:
+  name: clickhouse
+  namespace: {{ .Release.namespace }}
+  labels:
+    release: {{ .Values.clickhouse.monitoring.monitoringReleaseName }}
+spec:
+  endpoints:
+  - interval: 30s
+    port: clickhouse-metrics
+  - interval: 30s
+    port: operator-metrics
+  namespaceSelector:
+    matchNames:
+    - {{ .Release.namespace }}
+  selector:
+    matchLabels:
+      app: clickhouse-operator
+      argocd.argoproj.io/instance: {{ .Release.namespace }}
+      clickhouse.altinity.com/chop: 0.23.6
+{{- end }}
diff --git a/deployment/helm_chart/opik/templates/clickhouseinstallation.yaml b/deployment/helm_chart/opik/templates/clickhouseinstallation.yaml
@@ -23,6 +23,18 @@ spec:
       {{ .Values.clickhouse.adminUser.username }}/networks/ip:
         - '127.0.0.1/32'
         - '0.0.0.0/0'
+      {{- if .Values.clickhouse.monitoring.enables }}
+      {{- if .Values.clickhouse.monitoring.useSecret.enables }}
+      {{ .Values.clickhouse.monitoring.username }}/k8s_secret_password: {{ include "opik.name" .}}-{{ .Values.clickhouse.monitoring.secretname }}/{{ .Values.clickhouse.monitoring.password_key }}
+      {{- else }}
+      {{ .Values.clickhouse.monitoring.username }}/password: {{ .Values.clickhouse.monitoring.password }}
+      {{- end }}
+      {{- end }}
+    {{- if .Values.clickhouse.monitoring.enables }}
+    profiles:
+      {{ .Values.clickhouse.monitoring.username }}/readonly: 1
+    {{- end }}
+
     {{- if .Values.zookeeper.enabled }}
     zookeeper:
         nodes:

diff --git a/deployment/helm_chart/opik/values.yaml b/deployment/helm_chart/opik/values.yaml
@@ -216,6 +216,11 @@ clickhouse:
     # baseBackup: ""
   operator: 
     enabled: true
+  monitoring:
+    enabled: false
+    monitoringReleaseName: # if monitoring is enables, this must be set
+    username: opikmon # readonly user
+    password: opikmon
 
     # resources:
     #   requests: