From aecb15031db4d2e3e66df9f3aa10bd6abd5fd55b Mon Sep 17 00:00:00 2001
From: Mayne0213 <bluemayne0213@icloud.com>
Date: Thu, 8 Jan 2026 21:27:24 +0900
Subject: [PATCH] FEAT(grafana): add Thanos as default datasource

- Add Thanos Query as default Prometheus datasource
- Keep original Prometheus datasource as backup
- Thanos provides deduplicated metrics from HA Prometheus

REFACTOR(thanos): move all components to master node

- Add tolerations for control-plane:NoSchedule
- Add nodeSelector for control-plane node
- Affects: query, storegateway, compactor
- PVC will be recreated on master node (data in S3)

FIX(thanos): allow non-Bitnami images (quay.io/thanos)

FIX(thanos): correct nodeSelector value to 'true'
---
 grafana/helm-values.yaml | 10 +++++++++-
 thanos/helm-values.yaml  | 29 +++++++++++++++++++++++++++++
 2 files changed, 38 insertions(+), 1 deletion(-)

diff --git a/grafana/helm-values.yaml b/grafana/helm-values.yaml
index 2cf7e09..5eb65e2 100644
--- a/grafana/helm-values.yaml
+++ b/grafana/helm-values.yaml
@@ -54,11 +54,19 @@ datasources:
   datasources.yaml:
     apiVersion: 1
     datasources:
+    - name: Thanos
+      type: prometheus
+      access: proxy
+      url: http://thanos-query.thanos.svc.cluster.local:9090
+      isDefault: true
+      editable: true
+      jsonData:
+        timeInterval: "60s"
     - name: Prometheus
       type: prometheus
       access: proxy
       url: http://prometheus-kube-prometheus-prometheus.prometheus.svc.cluster.local:9090
-      isDefault: true
+      isDefault: false
       editable: true
     - name: Loki
       type: loki
diff --git a/thanos/helm-values.yaml b/thanos/helm-values.yaml
index 3e96530..229758f 100644
--- a/thanos/helm-values.yaml
+++ b/thanos/helm-values.yaml
@@ -7,6 +7,11 @@
 # - Store Gateway: reads historical data from MinIO
 # - Compactor: compacts and downsamples data in MinIO
 
+# Allow non-Bitnami images (quay.io/thanos/thanos)
+global:
+  security:
+    allowInsecureImages: true
+
 # Use quay.io image to avoid Docker Hub rate limits
 image:
   registry: quay.io
@@ -24,6 +29,14 @@ query:
   enabled: true
   replicaCount: 1
 
+  # Run on master node for stability
+  tolerations:
+    - key: node-role.kubernetes.io/control-plane
+      operator: Exists
+      effect: NoSchedule
+  nodeSelector:
+    node-role.kubernetes.io/control-plane: "true"
+
   # Deduplicate metrics from multiple Prometheus replicas
   dnsDiscovery:
     enabled: true
@@ -58,6 +71,14 @@ storegateway:
   enabled: true
   replicaCount: 1
 
+  # Run on master node for stability
+  tolerations:
+    - key: node-role.kubernetes.io/control-plane
+      operator: Exists
+      effect: NoSchedule
+  nodeSelector:
+    node-role.kubernetes.io/control-plane: "true"
+
   resources:
     requests:
       cpu: 15m
@@ -76,6 +97,14 @@ storegateway:
 compactor:
   enabled: true
 
+  # Run on master node for stability
+  tolerations:
+    - key: node-role.kubernetes.io/control-plane
+      operator: Exists
+      effect: NoSchedule
+  nodeSelector:
+    node-role.kubernetes.io/control-plane: "true"
+
   # Retention settings
   retentionResolutionRaw: 7d      # Keep raw data for 7 days
   retentionResolution5m: 30d      # Keep 5m downsampled for 30 days