FIX(velero): increase velero-ui memory limit

- Increase memory from 128Mi to 256Mi
- Fix OOMKilled (exit code 137) issue

cnpg/helm-values.yaml

@@ -8,15 +8,15 @@ image:
   pullPolicy: IfNotPresent
 
 # Replica count for operator
-replicaCount: 2
+replicaCount: 1
 
 # Resource requests for operator
 resources:
   requests:
     cpu: 15m
-    memory: 128Mi
+    memory: 121Mi
   limits:
-    memory: 128Mi
+    memory: 121Mi
 
 # RBAC
 rbac:
@@ -67,6 +67,9 @@ nodeSelector: {}
 # Tolerations
 tolerations: []
 
+# High priority for critical database operator
+priorityClassName: high-priority
+
 # Affinity - Soft Anti-Affinity to spread pods across nodes
 affinity:
   podAntiAffinity:

minio/helm-values.yaml

@@ -35,10 +35,10 @@ rootPasswordSecretKey: root-password
 # Resources
 resources:
   requests:
-    memory: 263Mi
+    memory: 175Mi
     cpu: 15m
   limits:
-    memory: 263Mi
+    memory: 175Mi
 
 # Service
 service:
@@ -58,7 +58,7 @@ environment:
   MINIO_API_REPLICATION_MAX_WORKERS: "1"
   MINIO_API_REPLICATION_MAX_LRG_WORKERS: "1"
   MINIO_API_TRANSITION_WORKERS: "1"
-  MINIO_SCANNER_SPEED: "slow"
+  MINIO_SCANNER_SPEED: "slowest"
 
 # API Ingress (S3 endpoint)
 ingress:
@@ -101,16 +101,18 @@ affinity:
                   - minio
           topologyKey: kubernetes.io/hostname
 
-# Tolerations for control-plane node (minio-0 runs on control-plane)
-tolerations:
-  - key: "node-role.kubernetes.io/control-plane"
-    operator: "Exists"
-    effect: "NoSchedule"
+# High priority for critical storage infrastructure
+priorityClassName: high-priority
 
 # Prometheus metrics
 metrics:
   serviceMonitor:
-    enabled: true  # Enable ServiceMonitor for Prometheus scraping
+    enabled: true
+    includeNode: true
+    additionalLabels:
+      release: prometheus
+    interval: 30s
+    scrapeTimeout: 10s
 
 # Disable post-install job by setting all triggers to empty
 # Job is created if any of: buckets, users, policies, customCommands, svcaccts exist

minio/manifests/console-deployment.yaml

@@ -6,7 +6,7 @@ metadata:
   labels:
     app: minio-console
 spec:
-  replicas: 2
+  replicas: 1
   selector:
     matchLabels:
       app: minio-console

pgweb/helm-values.yaml

@@ -3,7 +3,7 @@
 
 controllers:
   main:
-    replicas: 2
+    replicas: 1
     annotations:
       reloader.stakater.com/auto: "true"
     containers:

postgresql/manifests/cluster.yaml

@@ -67,10 +67,6 @@ spec:
   # Affinity to spread replicas across nodes (soft - prefer different nodes)
   affinity:
     podAntiAffinityType: preferred
-    tolerations:
-      - key: "node-role.kubernetes.io/control-plane"
-        operator: "Exists"
-        effect: "NoSchedule"
 
   # Enable superuser access
   enableSuperuserAccess: true

velero/argocd.yaml

@@ -74,9 +74,9 @@ spec:
         resources:
           requests:
             cpu: 15m  # VPA recommendation
-            memory: 128Mi
+            memory: 256Mi
           limits:
-            memory: 128Mi
+            memory: 256Mi
 
         service:
           type: ClusterIP
@@ -86,15 +86,6 @@ spec:
           - name: BASIC_AUTH_ENABLED
             value: "false"
 
-        # Run on master node with velero controller
-        nodeSelector:
-          node-role.kubernetes.io/control-plane: "true"
-
-        tolerations:
-          - key: "node-role.kubernetes.io/control-plane"
-            operator: "Exists"
-            effect: "NoSchedule"
-
         affinity: {}
         envFrom: []
 

velero/helm-values.yaml

@@ -17,9 +17,9 @@ replicaCount: 1
 resources:
   requests:
     cpu: 11m
-    memory: 128Mi
+    memory: 75Mi
   limits:
-    memory: 128Mi
+    memory: 75Mi
 
 # Init containers for plugins
 initContainers:
@@ -39,16 +39,10 @@ nodeAgent:
   privileged: false
   resources:
     requests:
-      cpu: 15m  # VPA recommendation: 15m
-      memory: 256Mi
+      cpu: 15m
+      memory: 100Mi
     limits:
-      memory: 256Mi
-  # Run on all nodes including control-plane for full backup coverage
-  tolerations:
-    - key: "node-role.kubernetes.io/control-plane"
-      operator: "Exists"
-      effect: "NoSchedule"
-
+      memory: 100Mi
 # Configuration for backup storage
 configuration:
   # Use existing BackupStorageLocation and VolumeSnapshotLocation
@@ -81,7 +75,8 @@ configuration:
   # This prevents status updates that trigger ArgoCD refreshes
   storeValidationFrequency: 0s
 
-  # Note: backup-sync is enabled to show backups in velero-ui
+  # Disable backup-sync (user only uses manual backups)
+  backupSyncPeriod: 0s
 
 # Credentials for S3 access (from Vault via External Secrets)
 credentials:
@@ -106,15 +101,5 @@ serviceAccount:
 rbac:
   create: true
 
-# Node selector - Run on control-plane node for stability
-nodeSelector:
-  node-role.kubernetes.io/control-plane: "true"
-
-# Tolerations - Allow scheduling on control-plane node
-tolerations:
-  - key: "node-role.kubernetes.io/control-plane"
-    operator: "Exists"
-    effect: "NoSchedule"
-
 # Affinity (disabled - single replica on master)
 affinity: {}

zot/helm-values.yaml

@@ -35,9 +35,9 @@ controllers:
         resources:
           requests:
             cpu: 15m
-            memory: 128Mi
+            memory: 121Mi
           limits:
-            memory: 128Mi
+            memory: 121Mi
 service:
   zot:
     controller: zot