weitere implementierung

README.md

@@ -0,0 +1,7 @@
+# Recovery
+
+⚠️ Restore NICHT über Argo deployen
+
+Use:
+
+kubectl apply -f restore-latest.yaml

argocd/applications/cnpg-operator.yaml

@@ -1,9 +1,10 @@
-# 1. App für den CloudNativePG Operator
 apiVersion: argoproj.io/v1alpha1
 kind: Application
 metadata:
   name: cnpg-operator
   namespace: argocd
+  annotations:
+    argocd.argoproj.io/sync-wave: "1"
 spec:
   project: default
   source:
@@ -19,10 +20,11 @@ spec:
   syncPolicy:
     automated:
       prune: true
-      selfHeel: true
+      selfHeal: true
     syncOptions:
     - CreateNamespace=true
     - ServerSideApply=true
+    - SkipDryRunOnMissingResource=true
     automated:
       allowEmpty: true
   ignoreDifferences:
@@ -30,5 +32,3 @@ spec:
     kind: CustomResourceDefinition
     jqPathExpressions:
     - .spec.conversion.webhook.clientConfig.caBundle
-  dependsOn:
-  - name: timescaledb-crds

argocd/applications/minio.yaml

@@ -27,7 +27,7 @@ spec:
   syncPolicy:
     automated:
       prune: true
-      selfHeel: true
+      selfHeal: true
     syncOptions:
     - CreateNamespace=true
     - ServerSideApply=true

argocd/applications/timescaledb-crds.yaml

@@ -3,6 +3,8 @@ kind: Application
 metadata:
   name: timescaledb-crds
   namespace: argocd
+  annotations:
+    argocd.argoproj.io/sync-wave: "0"
 spec:
   project: default
   source:
@@ -15,4 +17,4 @@ spec:
   syncPolicy:
     automated:
       prune: true
-      selfHeel: true
+      selfHeal: true

argocd/applications/timescaledb.yaml

@@ -3,6 +3,8 @@ kind: Application
 metadata:
   name: timescaledb
   namespace: argocd
+  annotations:
+    argocd.argoproj.io/sync-wave: "2"
 spec:
   project: default
   source:
@@ -15,14 +17,13 @@ spec:
   syncPolicy:
     automated:
       prune: true
-      selfHeel: true
+      selfHeal: true
     syncOptions:
     - CreateNamespace=true
     - ServerSideApply=true
+    - SkipDryRunOnMissingResource=true
   ignoreDifferences:
   - group: postgresql.cnpg.io
     kind: Cluster
     jqPathExpressions:
     - .status
-  dependsOn:
-  - name: cnpg-operator

argocd/root.yaml

@@ -0,0 +1,18 @@
+apiVersion: argoproj.io/v1alpha1
+kind: Application
+metadata:
+  name: root-apps
+  namespace: argocd
+spec:
+  project: default
+  source:
+    repoURL: https://gitea.vhn-demo.duckdns.org/hoang/timescaledb.git
+    path: argocd/applications
+    targetRevision: main
+  destination:
+    server: https://kubernetes.default.svc
+    namespace: argocd
+  syncPolicy:
+    automated:
+      prune: true
+      selfHeal: true

k8s/base/database/backup-daily.yaml

@@ -0,0 +1,11 @@
+apiVersion: postgresql.cnpg.io/v1
+kind: ScheduledBackup
+metadata:
+  name: timescale-backup-daily
+  namespace: database
+spec:
+  schedule: "0 2 * * *"   # täglich 02:00
+  backupOwnerReference: self
+  cluster:
+    name: timescale-cluster
+  method: barmanObjectStore

k8s/base/database/backup-hourly.yaml

@@ -0,0 +1,12 @@
+apiVersion: postgresql.cnpg.io/v1
+kind: ScheduledBackup
+metadata:
+  name: timescale-backup-hourly
+  namespace: database
+spec:
+  schedule: "0 * * * *"   # jede Stunden
+  backupOwnerReference: self
+  cluster:
+    name: timescale-cluster
+  method: barmanObjectStore
+

k8s/base/database/backup-monitor.yaml

@@ -0,0 +1,17 @@
+apiVersion: monitoring.coreos.com/v1
+kind: PrometheusRule
+metadata:
+  name: cnpg-backup-alert
+  namespace: database
+spec:
+  groups:
+    - name: cnpg.rules
+      rules:
+        - alert: CNPGBackupMissing
+          expr: time() - cnpg_last_backup_time_seconds > 86400
+          for: 10m
+          labels:
+            severity: critical
+          annotations:
+            summary: "CNPG Backup missing"
+            description: "No backup in last 24h"

k8s/base/database/cluster.yaml

@@ -4,10 +4,51 @@ metadata:
   name: timescale-cluster
 spec:
   instances: 3
+
   imageName: timescale/timescaledb:2.15.3-pg15
+
+  # ✅ Initiales Setup
+  bootstrap:
+    initdb:
+      database: app
+      owner: app
+
+  # ✅ Storage
   storage:
     size: 50Gi
+
+  # ✅ Ressourcen (wichtig!)
+  resources:
+    requests:
+      memory: "2Gi"
+      cpu: "500m"
+    limits:
+      memory: "4Gi"
+      cpu: "2"
+
+  # ✅ High Availability
+  affinity:
+    enablePodAntiAffinity: true
+    topologyKey: kubernetes.io/hostname
+
+  # ✅ Monitoring
+  monitoring:
+    enablePodMonitor: true
+
+  # ✅ PostgreSQL / Timescale Tuning
+  postgresql:
+    parameters:
+      max_connections: "100"        # mehr Verbindungen 200
+      shared_buffers: "256MB"       # Cache im RAM 1GB
+      effective_cache_size: "3GB"   # Optimizer hint
+      work_mem: "16MB"              # pro Query
+      maintenance_work_mem: "256MB" # für VACUUM/REINDEX
+      wal_buffers: "16MB"
+      checkpoint_completion_target: "0.9"
+      random_page_cost: "1.1"       # SSD optimiert
+      effective_io_concurrency: "200"
   backup:
+    retentionPolicy: "7d"   # Backup älter als 7 Tage -> automatisch gelöscht
     barmanObjectStore:
       destinationPath: "s3://backups/"
       endpointURL: "http://minio.minio-system.svc.cluster.local:9000"
@@ -18,3 +59,5 @@ spec:
         secretAccessKey:
           name: s3-creds
           key: SECRET_ACCESS_KEY
+      wal:
+        compression: gzip

k8s/base/database/kustomization.yaml

@@ -1,2 +1,9 @@
 resources:
   - cluster.yaml
+  - s3-secret.yaml
+  - backup-hourly.yaml
+  - backup-daily.yaml
+  - restore-cronjob.yaml
+  - restore-configmap.yaml
+
+

k8s/base/database/restore-configmap.yaml

@@ -0,0 +1,33 @@
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: restore-test-manifest
+  namespace: database
+data:
+  restore-test.yaml: |
+    apiVersion: postgresql.cnpg.io/v1
+    kind: Cluster
+    metadata:
+      name: timescale-restore-test
+    spec:
+      instances: 1
+      imageName: timescale/timescaledb:2.15.3-pg15
+      bootstrap:
+        recovery:
+          source: timescale-cluster
+          recoveryTarget:
+            targetImmediate: true
+      storage:
+        size: 10Gi
+      externalClusters:
+        - name: timescale-cluster
+          barmanObjectStore:
+            destinationPath: "s3://backups/"
+            endpointURL: "http://minio.minio-system.svc.cluster.local:9000"
+            s3Credentials:
+              accessKeyId:
+                name: s3-creds
+                key: S3_ACCESS_KEY_ID
+              secretAccessKey:
+                name: s3-creds
+                key: S3_SECRET_ACCESS_KEY

k8s/base/database/restore-cronjob.yaml

@@ -0,0 +1,34 @@
+apiVersion: batch/v1
+kind: CronJob
+metadata:
+  name: timescale-restore-test
+  namespace: database
+spec:
+  schedule: "0 4 * * *"   # täglich 04:00
+  jobTemplate:
+    spec:
+      template:
+        spec:
+          restartPolicy: OnFailure
+          containers:
+            - name: restore-test
+              image: bitnami/kubectl:latest
+              command:
+                - /bin/sh
+                - -c
+                - |
+                  kubectl apply -f /manifests/restore-test.yaml
+                  sleep 300
+
+                  kubectl get pods -n database
+
+                  echo "✅ Restore Test executed"
+
+                  kubectl delete cluster timescale-restore-test -n database || true
+              volumeMounts:
+                - name: manifests
+                  mountPath: /manifests
+          volumes:
+            - name: manifests
+              configMap:
+                name: restore-test-manifest

k8s/base/database/s3-secret.yaml

@@ -0,0 +1,9 @@
+apiVersion: v1
+kind: Secret
+metadata:
+  name: s3-creds
+  namespace: database
+type: Opaque
+stringData:
+  S3_ACCESS_KEY_ID: admin
+  S3_SECRET_ACCESS_KEY: password123

k8s/recovery/restore-latest.yaml

@@ -0,0 +1,25 @@
+apiVersion: postgresql.cnpg.io/v1
+kind: Cluster
+metadata:
+  name: timescale-restore-latest
+  namespace: database
+spec:
+  instances: 1
+  imageName: timescale/timescaledb:2.15.3-pg15
+  bootstrap:
+    recovery:
+      source: timescale-cluster
+  storage:
+    size: 50Gi
+  externalClusters:
+    - name: timescale-cluster
+      barmanObjectStore:
+        destinationPath: "s3://backups/"
+        endpointURL: "http://minio.minio-system.svc.cluster.local:9000"
+        s3Credentials:
+          accessKeyId:
+            name: s3-creds
+            key: S3_ACCESS_KEY_ID
+          secretAccessKey:
+            name: s3-creds
+            key: S3_SECRET_ACCESS_KEY

k8s/recovery/restore-test.yaml

@@ -0,0 +1,34 @@
+apiVersion: postgresql.cnpg.io/v1
+kind: Cluster
+metadata:
+  name: timescale-restore-test
+  namespace: database
+spec:
+  instances: 1
+  imageName: timescale/timescaledb:2.15.3-pg15
+  bootstrap:
+    recovery:
+      source: timescale-cluster
+      recoveryTarget:
+        targetImmediate: true   # sofort letzter WAL Stand
+  storage:
+    size: 10Gi   # kleiner für Tests
+  resources:
+    requests:
+      memory: "512Mi"
+      cpu: "200m"
+    limits:
+      memory: "1Gi"
+      cpu: "500m"
+  externalClusters:
+    - name: timescale-cluster
+      barmanObjectStore:
+        destinationPath: "s3://backups/"
+        endpointURL: "http://minio.minio-system.svc.cluster.local:9000"
+        s3Credentials:
+          accessKeyId:
+            name: s3-creds
+            key: S3_ACCESS_KEY_ID
+          secretAccessKey:
+            name: s3-creds
+            key: S3_SECRET_ACCESS_KEY

k8s/recovery/restore-timestamp.yaml

@@ -0,0 +1,27 @@
+apiVersion: postgresql.cnpg.io/v1
+kind: Cluster
+metadata:
+  name: timescale-restore-pitr
+  namespace: database
+spec:
+  instances: 1
+  imageName: timescale/timescaledb:2.15.3-pg15
+  bootstrap:
+    recovery:
+      source: timescale-cluster
+      recoveryTarget:
+        targetTime: "2026-05-06 02:30:00"   # ANPASSEN
+  storage:
+    size: 50Gi
+  externalClusters:
+    - name: timescale-cluster
+      barmanObjectStore:
+        destinationPath: "s3://backups/"
+        endpointURL: "http://minio.minio-system.svc.cluster.local:9000"
+        s3Credentials:
+          accessKeyId:
+            name: s3-creds
+            key: S3_ACCESS_KEY_ID
+          secretAccessKey:
+            name: s3-creds
+            key: S3_SECRET_ACCESS_KEY

scripts/restore.sh

@@ -0,0 +1,32 @@
+#!/bin/bash
+
+set -e
+
+MODE=$1
+TIMESTAMP=$2
+
+echo "🚀 Starting restore..."
+
+if [ "$MODE" == "latest" ]; then
+  kubectl apply -f restore-latest.yaml
+
+elif [ "$MODE" == "pitr" ]; then
+  if [ -z "$TIMESTAMP" ]; then
+    echo "❌ Please provide timestamp"
+    exit 1
+  fi
+
+  sed "s/TIMESTAMP/$TIMESTAMP/" restore-timestamp.yaml | kubectl apply -f -
+
+elif [ "$MODE" == "test" ]; then
+  kubectl apply -f restore-test.yaml
+
+else
+  echo "Usage:"
+  echo "  ./restore.sh latest"
+  echo "  ./restore.sh pitr '2026-05-06 02:30:00'"
+  echo "  ./restore.sh test"
+  exit 1
+fi
+
+echo "✅ Restore triggered"