Cloudera Upgrade and Migrations Paths

Introduction to Cloudera🔗

The merger of Cloudera and Hortonworks led to the new Cloudera, which is the combined best of breed Big Data components from both Cloudera and Hortonworks.

Review the following information before you upgrade or migrate to Cloudera:

Cloudera Overview
Cloudera Base on premises new features

Troubleshooting🔗

A selection of Cloudera Knowledge Base articles are available that describe common issues encountered by Cloudera customers during upgrades and migrations. See Cloudera Upgrade/Migrate Troubleshooting Articles. (Cloudera login required.)

Cloudera upgrade and migration paths are: 🔗

Cloudera Upgrade and Migration paths for Data-at-Rest
Cloudera Upgrade and Migration paths to Cloudera for Cloudera DataFlow Functions components
Cloudera Upgrade and Migration paths for Cloudera Data Science Workbench (Cloudera AI)

Cloudera Upgrade and Migration Paths for Data-at-Rest🔗

If you are an HDP or a CDH user, you can follow one of the several upgrade or migration paths to Cloudera.

Migration and Upgrade Paths	Description
In-place upgrade	Recommended for large clusters. Other paths are not viable. Involves downtime. See the following to know more about the In-place upgrade paths: In-place upgrade from CDH to CDP Private Cloud Base In-place upgrade from HDP to CDP Private Cloud Base
Migrate to new cluster	Recommended if you are ready for a hardware refresh or have small clusters. You can fall back to your original cluster in the event of upgrade issues. Requires additional hardware.
Direct to cloud migration	Recommended if you can tolerate some cluster downtime or have bursty workloads. See the following to know more about the direct to cloud migration paths: Nifi workloads If you are running NiFi workloads without Hive, Impala, HBase, or Kafka: Migrate to Cloudera Data Hub and use the Flow Management cluster template. Kafka workloads If you are running Kafka workloads without Hive, Impala, HBase, or NiFi: Migrate to Cloudera Data Hub and use the Streams Messaging cluster template. HBase workloads If you are running HBase workloads without Hive or Impala: Migrate to Cloudera Data Hub and use the Cloudera Operational Database cluster template. Hive or Impala workloads If you are running Hive or Impala workloads without HBase: Migrate to Cloudera Data Warehouse. Other workloads Migrate to Cloudera Data Hub and use the custom cluster template.

Cloudera Upgrade and Migration Paths to Cloudera for Cloudera DataFlow Functions components🔗

You can upgrade or migrate Cloudera DataFlow Functions components to Cloudera in the following ways:

What do you want to migrate?	Do you want to migrate or upgrade?
Streaming workloads	In-place upgrade from CDH In-place upgrade from HDP Migrating Streaming workloads from HDF to Cloudera Base on premises
Flow Management workloads	Follow the applicable upgrade or migration path in Upgrade and migration paths.
Workloads for deprecated components	Migrate Flume workloads to NiFi for Cloudera. This workflow is in development. Migrate Storm workloads to Flink for Cloudera. This content is in development.

Cloudera Migration and Upgrade Paths for Cloudera Data Science Workbench (Cloudera AI)🔗

You can upgrade or migrate Cloudera Data Science Workbench to Cloudera AI in the following ways:

What do you want to migrate?	Do you want to migrate or upgrade?
Cloudera Data Science Workbench with CDH/HDP to Cloudera Data Science Workbench with Cloudera Base on premises	For in-place upgrade follow the upgrade documentation. For migration follow the documented migration steps.
Cloudera Data Science Workbench to Cloudera AI-Cloudera on cloud	The workflow and tools to migrate from Cloudera Data Science Workbench to Cloudera AI in Cloudera on cloud are in development. To perform this migration now, contact your Cloud account or professional services representative.
Cloudera Data Science Workbench to Cloudera AI-Cloudera Base on premises	The workflow and tools to migrate from Cloudera Data Science Workbench to Cloudera AI in Cloudera Base on premises is available as technical preview. For more information, see the migration documentation.