将 HDInsight 群集迁移到较新版本Migrate HDInsight cluster to a newer version

若要利用最新的 HDInsight 功能,建议将 HDInsight 群集定期迁移到最新版本。To take advantage of the latest HDInsight features, we recommend that HDInsight clusters be regularly migrated to latest version. HDInsight 不支持将现有群集升级到较新组件版本的就地升级。HDInsight does not support in-place upgrades where an existing cluster is upgraded to a newer component version. 必须使用所需的组件和平台版本创建新群集,然后迁移应用程序以使用新群集。You must create a new cluster with the desired component and platform version and then migrate your applications to use the new cluster. 按照以下准则迁移 HDInsight 群集版本。Follow the below guidelines to migrate your HDInsight cluster versions.

备注

有关支持的 HDInsight 版本的信息,请参阅 HDInsight 组件版本For information on supported versions of HDInsight, see HDInsight component versions.

迁移任务Migration tasks

升级 HDInsight 群集的工作流如下所示。The workflow to upgrade HDInsight Cluster is as follows.

升级工作流示意图

  1. 请阅读本文档的每个部分,了解在升级 HDInsight 群集时可能需要进行的更改。Read each section of this document to understand changes that may be required when upgrading your HDInsight cluster.
  2. 创建群集作为测试/质量保证环境。Create a cluster as a test/quality assurance environment. 有关创建群集的详细信息,请参阅了解如何创建基于 Linux 的 HDInsight 群集For more information on creating a cluster, see Learn how to create Linux-based HDInsight clusters
  3. 将现有作业、数据源及接收器复制到新环境。Copy existing jobs, data sources, and sinks to the new environment.
  4. 执行验证测试,以确保作业在新群集上按预期工作。Perform validation testing to make sure that your jobs work as expected on the new cluster.

验证一切都按预期工作后,请为迁移安排停机时间。Once you have verified that everything works as expected, schedule downtime for the migration. 在此停机期间,请执行以下操作:During this downtime, do the following actions:

  1. 备份所有存储在本地群集节点上的暂时性数据。Back up any transient data stored locally on the cluster nodes. 例如,如果数据直接存储在头节点上。For example, if you have data stored directly on a head node.
  2. 删除现有群集Delete the existing cluster.
  3. 使用前一群集使用的同一默认数据存储在具有最新(或支持)的 HDI 版本的同一 VNET 子网中创建群集。Create a cluster in the same VNET subnet with latest (or supported) HDI version using the same default data store that the previous cluster used. 这样,新群集便可针对现有生产数据继续运行。This allows the new cluster to continue working against your existing production data.
  4. 导入任何已备份的暂时性数据。Import any transient data you backed up.
  5. 使用新群集启动作业/继续处理。Start jobs/continue processing using the new cluster.

特定于工作负载的指南Workload specific guidance

以下文档提供有关如何迁移特定工作负载的指南:The following documents provide guidance on how to migrate specific workloads:

备份和还原Backup and restore

有关数据库备份和还原的详细信息,请参阅使用自动数据库备份在 Azure SQL 数据库中恢复数据库For more information about database backup and restore, see Recover a database in Azure SQL Database by using automated database backups.

后续步骤Next steps