Upgrade to 0.9.0 (Thakurvaibhav#12)

Zahidmad · Dec 11, 2019 · 1235339 · 1235339
1 parent b880e05
commit 1235339
Show file tree

Hide file tree

Showing 5 changed files with 45 additions and 20 deletions.
diff --git a/monitoring/prometheus-ha/01-prometheus-configmap.yaml b/monitoring/prometheus-ha/01-prometheus-configmap.yaml
@@ -13,7 +13,7 @@ data:
       external_labels:
         cluster: prometheus-ha
         # Each Prometheus has to have unique labels.
-        replica: $(POD_NAME)
+        prometheus_replica: $(POD_NAME)
 
     rule_files:
       - /etc/prometheus/rules/*rules.yaml

diff --git a/monitoring/prometheus-ha/05-thanos-querier.yaml b/monitoring/prometheus-ha/05-thanos-querier.yaml
@@ -17,15 +17,17 @@ spec:
     spec:
       containers:
       - name: thanos
-        image: quay.io/thanos/thanos:v0.8.0
+        image: quay.io/thanos/thanos:v0.9.0
         args:
         - query
         - --log.level=debug
         - --query.replica-label=replica
+        - --grpc-address=0.0.0.0:10901
+        - --http-address=0.0.0.0:9090
         - --store=dnssrv+thanos-store-gateway:10901
         ports:
         - name: http
-          containerPort: 10902
+          containerPort: 9090
         - name: grpc
           containerPort: 10901
         livenessProbe:
@@ -36,6 +38,13 @@ spec:
           httpGet:
             port: http
             path: /-/ready
+        resources:
+          limits:
+            cpu: "1"
+            memory: 1Gi
+          requests:
+            cpu: 100m
+            memory: 256Mi
 ---
 apiVersion: v1
 kind: Service

diff --git a/monitoring/prometheus-ha/06-thanos-ruler.yaml b/monitoring/prometheus-ha/06-thanos-ruler.yaml
@@ -38,18 +38,21 @@ spec:
     spec:
       containers:
         - name: thanos
-          image: quay.io/thanos/thanos:v0.8.0
+          image: quay.io/thanos/thanos:v0.9.0
           args:
             - rule
+            - --grpc-address=0.0.0.0:10901
+            - --http-address=0.0.0.0:10902
             - --log.level=debug
             - --data-dir=/data
-            - --eval-interval=15s
+            - --eval-interval=10s
             - --rule-file=/etc/thanos-ruler/*.rules.yaml
             - --alertmanagers.url=http://alertmanager:9093
             - --query=thanos-querier:9090
             - "--objstore.config={type: GCS, config: {bucket: thanos-ruler}}"
             - --label=ruler_cluster="prometheus-ha"
             - --label=replica="$(POD_NAME)"
+            - --alert.label-drop="replica"
           env:
             - name : GOOGLE_APPLICATION_CREDENTIALS
               value: /etc/secret/thanos-gcs-credentials.json
@@ -63,13 +66,26 @@ spec:
             - name: grpc
               containerPort: 10901
           livenessProbe:
+            failureThreshold: 4
             httpGet:
-              port: http
               path: /-/healthy
+              port: 10902
+              scheme: HTTP
+            periodSeconds: 30
           readinessProbe:
             httpGet:
-              port: http
               path: /-/ready
+              port: 10902
+              scheme: HTTP
+            initialDelaySeconds: 10
+            periodSeconds: 30
+          resources:
+            limits:
+              cpu: "1"
+              memory: 1Gi
+            requests:
+              cpu: 50m
+              memory: 100Mi
           volumeMounts:
             - mountPath: /etc/thanos-ruler
               name: config

diff --git a/monitoring/prometheus-ha/07-thanos-store-gateway.yaml b/monitoring/prometheus-ha/07-thanos-store-gateway.yaml
@@ -19,14 +19,14 @@ spec:
     spec:
       containers:
         - name: thanos
-          image: quay.io/thanos/thanos:v0.8.0
+          image: quay.io/thanos/thanos:v0.9.0
           args:
           - "store"
+          - --grpc-address=0.0.0.0:10901
+          - --http-address=0.0.0.0:10902
           - "--log.level=debug"
           - "--data-dir=/data"
           - "--objstore.config={type: GCS, config: {bucket: prometheus-long-term}}"
-          - "--index-cache-size=500MB"
-          - "--chunk-pool-size=500MB"
           env:
             - name : GOOGLE_APPLICATION_CREDENTIALS
               value: /etc/secret/thanos-gcs-credentials.json
@@ -38,18 +38,19 @@ spec:
           resources:
             limits:
               cpu: 1
-              memory: 5Gi
+              memory: 8Gi
             requests:
               cpu: 0.5
               memory: 1Gi
-          livenessProbe:
-            httpGet:
-              port: 10902
-              path: /-/healthy
           readinessProbe:
+            failureThreshold: 6
             httpGet:
               port: 10902
               path: /-/ready
+            periodSeconds: 20
+            successThreshold: 1
+            timeoutSeconds: 5
+          terminationMessagePolicy: FallbackToLogsOnError
           volumeMounts:
             - name: thanos-gcs-credentials
               mountPath: /etc/secret

diff --git a/monitoring/prometheus-ha/08-thanos-compacter.yaml b/monitoring/prometheus-ha/08-thanos-compacter.yaml
@@ -18,11 +18,14 @@ spec:
     spec:
       containers:
         - name: thanos
-          image: quay.io/thanos/thanos:v0.8.0
+          image: quay.io/thanos/thanos:v0.9.0
           args:
             - "compact"
             - "--log.level=debug"
             - "--data-dir=/data"
+            - --retention.resolution-raw=15d
+            - --retention.resolution-5m=10d
+            - --retention.resolution-1h=10d
             - "--objstore.config={type: GCS, config: {bucket: prometheus-long-term}}"
             - "--wait"
           env:
@@ -31,10 +34,6 @@ spec:
           ports:
             - name: http
               containerPort: 10902
-          livenessProbe:
-            httpGet:
-              port: 10902
-              path: /-/healthy
           readinessProbe:
             httpGet:
               port: 10902