dendrite-v16.yaml

---
apiVersion: postgresql.cnpg.io/v1
kind: Cluster
metadata:
  name: dendrite-psql-v16
  namespace: default
  annotations:
    kyverno.io/ignore: "true"
spec:
  # using only 1 replica: very difficult to drain the node where postgres is running
  # using more than 1 replica: write amplification issues when leveraging replicated storage (e.g. ceph)
  instances: 2
  imageName: ghcr.io/cloudnative-pg/postgresql:16.4
  primaryUpdateStrategy: unsupervised
  storage:
    size: 25Gi
    storageClass: rook-ceph-block
  superuserSecret:
    name: cloudnative-pg
  postgresql:
    parameters:
      max_connections: "300"
      shared_buffers: 512MB
  monitoring:
    enablePodMonitor: true
  resources:
    requests:
      memory: "512Mi"
    limits:
      memory: "2Gi"
  affinity:
    tolerations:
    - key: "arm"
      operator: "Exists"
  backup:
    retentionPolicy: 30d
    barmanObjectStore:
      wal:
        compression: bzip2
        maxParallel: 4
      destinationPath: s3://postgresql/
      endpointURL: http://minio.kube-system.svc:9000
      serverName: dendrite-v16
      s3Credentials:
        accessKeyId:
          name: cloudnative-pg
          key: MINIO_ACCESS_KEY
        secretAccessKey:
          name: cloudnative-pg
          key: MINIO_SECRET_KEY
  # bootstrap:
  #   recovery:
  #     source: dendrite-psql-backup
  # externalClusters:
  #   - name: dendrite-psql-backup
  #     barmanObjectStore:
  #       wal:
  #         compression: bzip2
  #         maxParallel: 4
  #       destinationPath: s3://postgresql/
  #       endpointURL: http://minio.kube-system.svc:9000
  #       serverName: dendrite-v16
  #       s3Credentials:
  #         accessKeyId:
  #           name: cloudnative-pg
  #           key: MINIO_ACCESS_KEY
  #         secretAccessKey:
  #           name: cloudnative-pg
  #           key: MINIO_SECRET_KEY