Upgrading a Cluster (Parcel)
Upgrading a cluster can entail upgrading the Hadoop distribution, upgrading Pepperdata, or upgrading both Hadoop and Pepperdata. For supported Pepperdata upgrade paths, you can skip Pepperdata versions without having to perform intermediate upgrades, and you can rollback a Pepperdata upgrade to your previously installed version. In accordance with best practices for upgrades, we recommend that you first upgrade a test cluster and verify successful operation before you upgrade a production cluster.
If you want to perform an upgrade in a CDP Public Cloud environment, you must create a new environment and Data Hub cluster, and install the Pepperdata Supervisor version that you want; see Installing Pepperdata (CDP Public Cloud).
When to Upgrade Pepperdata
To decide whether to upgrade Pepperdata, consider whether:
-
There are new features that you’d like to use.
-
There are bug fixes that are applicable to your environment. (See Pepperdata Release Notes.)
-
You want to upgrade your cluster’s Hadoop distribution to a version that is not supported by your installed Pepperdata version but that a newer Pepperdata version supports. (See Pepperdata-Platform Support.)
-
You need assistance, but your version of Pepperdata is no longer supported. (For the list of supported versions, see Pepperdata-Platform Support.)
Upgrade and Rollback Procedures
For upgrade and rollback procedures, see the following: