completely reworked

argo-git.yaml

@@ -0,0 +1,21 @@
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: oadp-backups
+  namespace: openshift-gitops
+spec:
+  project: default
+  source:
+    repoURL: <YOUR_REPO_URL>
+    targetRevision: main
+    path: gitops/oadp/base
+  destination:
+    server: https://kubernetes.default.svc
+    namespace: openshift-adp
+  syncPolicy:
+    automated:
+      prune: true
+      selfHeal: true
+    syncOptions:
+      - CreateNamespace=false
+      - ServerSideApply=true

backup-test-cronjob.yaml

@@ -0,0 +1,63 @@
+---
+# Optional: Monthly restore test automation
+apiVersion: batch/v1
+kind: CronJob
+metadata:
+  name: monthly-restore-test
+  namespace: openshift-adp
+spec:
+  schedule: "0 06 15 * *"  # 15th of month, 6 AM
+  concurrencyPolicy: Forbid
+  successfulJobsHistoryLimit: 3
+  failedJobsHistoryLimit: 3
+  jobTemplate:
+    spec:
+      template:
+        metadata:
+          labels:
+            app: restore-test
+        spec:
+          serviceAccountName: velero
+          restartPolicy: OnFailure
+          containers:
+          - name: restore-test
+            image: quay.io/konveyor/velero:latest
+            env:
+              - name: VELERO_NAMESPACE
+                value: openshift-adp
+            command:
+              - /bin/bash
+              - -c
+              - |
+                set -e
+                
+                echo "=== Velero Restore Test ==="
+                echo "Date: $(date)"
+                
+                # Get latest daily-config backup
+                CONFIG_BACKUP=$(velero backup get --selector="backup-type=config" \
+                  -o json | jq -r '.items[0].metadata.name')
+                
+                # Get latest daily-stateful backup  
+                STATEFUL_BACKUP=$(velero backup get --selector="backup-type=stateful" \
+                  -o json | jq -r '.items[0].metadata.name')
+                
+                echo "Latest config backup: $CONFIG_BACKUP"
+                echo "Latest stateful backup: $STATEFUL_BACKUP"
+                
+                # Verify backups are successful
+                CONFIG_STATUS=$(velero backup get $CONFIG_BACKUP -o json | \
+                  jq -r '.status.phase')
+                STATEFUL_STATUS=$(velero backup get $STATEFUL_BACKUP -o json | \
+                  jq -r '.status.phase')
+                
+                echo "Config backup status: $CONFIG_STATUS"
+                echo "Stateful backup status: $STATEFUL_STATUS"
+                
+                if [ "$CONFIG_STATUS" != "Completed" ] || [ "$STATEFUL_STATUS" != "Completed" ]; then
+                  echo "ERROR: Backups not in Completed state"
+                  exit 1
+                fi
+                
+                echo "=== Test Passed ==="
+                echo "All backups verified successfully"

config-backups.yaml

@@ -1,20 +1,41 @@
+---
+# Schedule 1: Daily config-only backup (fast, all namespaces)
 apiVersion: velero.io/v1
 kind: Schedule
 metadata:
   name: daily-config
   namespace: openshift-adp
 spec:
-  schedule: "0 02 * * *"
-
+  schedule: "0 02 * * *"  # 2 AM daily
+  
   # Make backups readable, sortable, unique
   nameTemplate: "{{ .ScheduleName }}-{{ .Timestamp }}"
 
   template:
+    # Backup all namespaces
     includedNamespaces:
       - "*"
-
-    # No PV data — just manifests
+    
+    # Labels for filtering and reporting
+    labels:
+      backup-type: config
+      schedule: daily
+      retention: short
+    
+    # Exclude ephemeral/generated resources
+    excludedResources:
+      - events
+      - events.events.k8s.io
+      - pipelineruns.tekton.dev    # Completed pipeline runs
+      - taskruns.tekton.dev          # Completed task runs
+      - replicasets.apps             # Managed by deployments
+      - pods                         # Recreated by controllers
+      - endpoints                    # Auto-generated
+      - endpointslices.discovery.k8s.io
+    
+    # No volume data - manifests only
     snapshotVolumes: false
     defaultVolumesToFsBackup: false
-
-    ttl: 336h
+    
+    # 14 days retention
+    ttl: 336h

dpa.yaml

@@ -1,3 +1,5 @@
+---
+# DPA Configuration - optimized for SNO with Kopia
 apiVersion: oadp.openshift.io/v1alpha1
 kind: DataProtectionApplication
 metadata:
@@ -28,6 +30,14 @@ spec:
     nodeAgent:
       enable: true
       uploaderType: kopia
+      podConfig:
+        resourceAllocations:
+          limits:
+            cpu: "1"        # Increased for database compression
+            memory: "1Gi"   # Increased for larger chunks
+          requests:
+            cpu: "200m"
+            memory: "512Mi"
     velero:
       defaultPlugins:
         - openshift
@@ -45,4 +55,4 @@ spec:
             cpu: "100m"
             memory: "256Mi"
 
-  logFormat: text
+  logFormat: text

stateful-backups.yaml

@@ -1,14 +1,17 @@
+---
+# Schedule 2: Daily stateful backup (with volume data)
 apiVersion: velero.io/v1
 kind: Schedule
 metadata:
   name: daily-stateful
   namespace: openshift-adp
 spec:
-  schedule: "0 03 * * *"
-
+  schedule: "0 03 * * *"  # 3 AM daily (after config backup)
+  
   nameTemplate: "{{ .ScheduleName }}-{{ .Timestamp }}"
 
   template:
+    # Only namespaces with persistent data
     includedNamespaces:
       - gitea
       - authentik
@@ -17,10 +20,80 @@ spec:
       - n8n
       - apim
       - gitea-ci
-
-
-    # No CSI snapshots; use nodeAgent/kopia only
+    
+    labels:
+      backup-type: stateful
+      schedule: daily
+      retention: short
+    
+    # Exclude ephemeral resources
+    excludedResources:
+      - events
+      - events.events.k8s.io
+      - pipelineruns.tekton.dev
+      - taskruns.tekton.dev
+      - replicasets.apps
+      - pods
+    
+    # Use Kopia for volume backups
     snapshotVolumes: false
     defaultVolumesToFsBackup: true
-
+    
+    # 14 days retention
     ttl: 336h
+    
+    # Pre-backup hooks for data consistency
+    hooks:
+      resources:
+        # Cassandra: flush memtables to disk before backup
+        - name: cassandra-flush
+          includedNamespaces:
+            - cassandra
+          labelSelector:
+            matchLabels:
+              app.kubernetes.io/name: cassandra
+          pre:
+            - exec:
+                container: cassandra
+                command:
+                  - /bin/bash
+                  - -c
+                  - nodetool flush
+                timeout: 5m
+                onError: Continue
+        
+        # Gitea PostgreSQL: checkpoint before backup
+        - name: gitea-postgres-checkpoint
+          includedNamespaces:
+            - gitea
+          labelSelector:
+            matchLabels:
+              app.kubernetes.io/name: postgresql
+              app.kubernetes.io/instance: gitea
+          pre:
+            - exec:
+                container: postgresql
+                command:
+                  - /bin/bash
+                  - -c
+                  - psql -U postgres -c 'CHECKPOINT;'
+                timeout: 2m
+                onError: Continue
+        
+        # Authentik PostgreSQL: checkpoint before backup
+        - name: authentik-postgres-checkpoint
+          includedNamespaces:
+            - authentik
+          labelSelector:
+            matchLabels:
+              app.kubernetes.io/name: postgresql
+              app.kubernetes.io/instance: authentik
+          pre:
+            - exec:
+                container: postgresql
+                command:
+                  - /bin/bash
+                  - -c
+                  - psql -U postgres -c 'CHECKPOINT;'
+                timeout: 2m
+                onError: Continue