Upgrading Database Catalogs and Virtual Warehouses in CDW Private Cloud

After you upgrade the CDP Private Cloud Data Services platform, you must upgrade the Database Catalog and Virtual Warehouses in Cloudera Data Warehouse (CDW). Upgrading to the latest release brings you new features from Hive, Impala, Hue, and other related runtime services. This is known as an in-place upgrade.

What gets upgraded

Database Catalog in CDW uses a Hive MetaStore (HMS) instance. The Virtual Warehouses use Apache Hive, Apache Impala, and Hue runtime images that are used in CDW. These runtime images are different than those used on CDP Private Cloud Base. With every new CDP Private Cloud Data Services release, you get a new version of Apache Hive, Apache Impala, and Hue runtimes with CDW, which includes new features and fixes.

Supported upgrade path for an in-place upgrade

In-place upgrade option is available only for upgrades from CDP Private Cloud Data Services 1.5.0 to a newer release.

What you should know before you upgrade

Review the Release Notes to learn about the new features, fixes, and known issues in this release, and more importantly, the upgrade-related known issues.

In-place upgrade steps

To perform an in-place upgrade:
  1. Upgrade the CDP Private Cloud Data Services platform.
  2. Log in to the Data Warehouse service as DWAdmin.
  3. Upgrade the Database Catalog by clicking > Upgrade.
  4. Upgrade individual Virtual Warehouses by clicking > Upgrade.

To verify a successful upgrade, check the version information on the Database Catalog or Virtual Warehouse details page by clicking > Edit on the Database Catalog or Virtual Warehouse tile.

What changes after the upgrade

  • If you had enabled the setting to copy the base cluster configurations to CDW on the CDP Private Cloud Data Services 1.5.1 cluster, then the base cluster configurations will continue to get copied to CDW after the upgrade. However, you can disable this setting from the Advanced Settings page.
  • Starting with CDP Private Cloud Data Services 1.5.1, Data Analytics Studio (DAS) has been deprecated and completely removed from CDW. After you upgrade the platform, any running DAS instances will be removed from the cluster. Cloudera recommends that you use Hue for querying and exploring data in CDW.
  • If you upgrade the CDP Platform from 1.5.0 to the latest release, then the configuration of an existing environments stays the same as before. Configurations such as default file format, compression type, and transactional type are not copied from the base cluster. To copy configurations from the base cluster, you must reactivate the environment.
  • Starting with CDP Private Cloud Data Services 1.5.0, Hue in CDW requires WebHDFS to be enabled on the CDP Private Cloud Base cluster. Ensure that worker nodes for both, OpenShift Container Platform (OCP) and Embedded Container Service (ECS), have access to the WebHDFS (HTTPFS) port 14000.