将 Cloud Pak for Data 卷备份到持久卷声明或对象库

cpd-cli backup-restore 下面的命令是一个示例，当您使用 S3 对象存储来存储备份，并且群集可以访问 icr.io/cpopen/cpd.

# Initialize the cpdbr first with pvc name and s3 storage.  Note that the bucket must exist.
# Example for cluster with access to ICR
$ cpd-cli backup-restore init \
--namespace $NAMESPACE \
--pvc-name cpdbr-pvc \
--image-prefix=icr.io/cpopen/cpd \
--provider=s3 \
--s3-endpoint="s3 endpoint" \
--s3-bucket=cpdbr \
--s3-prefix=$NAMESPACE/

以下命令是在使用专用映像注册表的环境中 (例如，集群气郄时) 使用 S3 对象存储库来存储备份时初始化 cpd-cli backup-restore 的示例。

# Example for air-gapped environment
$ cpd-cli backup-restore init \
--namespace $NAMESPACE \
--pvc-name cpdbr-pvc \
--image-prefix=${PRIVATE_REGISTRY_LOCATION}/ubi8 \
--provider=s3 \
--s3-endpoint="s3 endpoint" \
--s3-bucket=cpdbr \
--s3-prefix=$NAMESPACE/

cpd-cli backup-restore 以下命令是一个示例，用于在使用单独的 PVC 存储备份且群集可访问 icr.io/cpopen/cpd.

# Example for cluster with access to ICR
cpd-cli backup-restore init \
--namespace $NAMESPACE \
--log-level=debug \
--verbose \
--pvc-name cpdbr-pvc \ 
--image-prefix=icr.io/cpopen/cpd \
--provider=local

以下命令是在使用单独 PVC 来存储备份时初始化 cpd-cli backup-restore 的示例，例如，在使用专用映像注册表的环境中 (例如，当集群气郄时)。

# Example for air-gapped environment
cpd-cli backup-restore init \
--namespace $NAMESPACE \
--log-level=debug \
--verbose \
--pvc-name cpdbr-pvc \ 
--image-prefix=${PRIVATE_REGISTRY_LOCATION}/ubi8 \
--provider=local

仅当 EDB Postgres 是嵌入式 EDB Postgres 数据库时，才会对其进行备份。

export IMAGE_ZEN_UTILS=$(oc get cronjob zen-metastore-backup-cron-job -o=jsonpath='{.spec.jobTemplate.spec.template.spec.containers[?(@.name=="zen-metastore-backup-cron-job")].image}')

cat <<EOF | oc apply -f -
apiVersion: batch/v1
kind: Job
metadata:
  name: zen-edb-volumes-backup-job
  labels:
    app: 0020-zen-base
    component: zen-core-configmaps
    release: 0020-core
spec:
  template:
    metadata:
      labels:
        app: 0020-zen-base
        component: zen-core-configmaps
        release: 0020-core
      annotations:
        productMetric: FREE
        productName: IBM Cloud Platform Common Services
        productVersion: 5.0.0
    spec:
      serviceAccountName: zen-admin-sa
      containers:
      - name: backup-metastore-edb-data
        image: $IMAGE_ZEN_UTILS
        imagePullPolicy: IfNotPresent
        resources:
          limits:
            cpu: 500m
            ephemeral-storage: 512Mi
            memory: 2Gi
          requests:
            cpu: 100m
            ephemeral-storage: 128Mi
            memory: 512Mi
        command:
        - sh
        - -ce
        - |
            bash /tmp/zen-volume-scripts/cnpsql-backup-script.sh
        envFrom:
        - configMapRef:
            name: ibm-zen-metastore-edb-cm
        - configMapRef:
            name: ibm-zen-objectstore-cm
        env:
        - name: ICPD_CONTROLPLANE_NAMESPACE
          valueFrom:
            fieldRef:
              fieldPath: metadata.namespace
        volumeMounts:
        - mountPath: /user-home
          name: zen-objstore-local
        - name: metastore-edb-secret
          mountPath: /postgres/secrets
        - mountPath: /objectstore-backup
          name: ibm-zen-objectstore-backup-mount
          readOnly: false
        - mountPath: /tmp/zen-volume-scripts
          name: zen-volume-utils-cm
          readOnly: false
      volumes:
      - emptyDir: {}
        name: zen-objstore-local
      - name: metastore-edb-secret
        secret:
          secretName: ibm-zen-metastore-edb-secret
          defaultMode: 420
      - name: ibm-zen-objectstore-backup-mount
        persistentVolumeClaim:
          claimName: ibm-zen-objectstore-backup-pvc
      - name: zen-volume-utils-cm
        configMap:
          defaultMode: 511
          name: zen-volume-utils-cm
      restartPolicy: OnFailure
EOF

cpd-cli backup-restore volume-backup download <backup_name>

cpd-cli backup-restore volume-backup status <backup_name> -n ${PROJECT_CPD_INST_OPERANDS}

cpd-cli backup-restore volume-backup list -n ${PROJECT_CPD_INST_OPERANDS}

cpd-cli backup-restore volume-backup list --details -n ${PROJECT_CPD_INST_OPERANDS}

cpd-cli backup-restore volume-backup logs <backup_name> -n ${PROJECT_CPD_INST_OPERANDS}

cpd-cli backup-restore reset -n ${PROJECT_CPD_INST_OPERANDS} --force

<number_of_replicas> 的值取决于安装 Data Refinery 时的 scaleConfig 设置 (1 表示小型， 3 表示中型， 4 表示大型)。

oc scale --replicas=<number_of_replicas> deploy wdp-shaper wdp-dataprep