AWS and Azure Multi-AZ for Data Hubs

By default, Cloudera provisions Data Hubs in a single AWS or Azure availability zone (AZ), but you can optionally choose to deploy them across multiple availability zones (multi-AZ).

Data Hub clusters can be created from multiple different templates like “Data Engineering” and “Data Mart”. You can also create custom templates to deploy a custom set of supported components such as Yarn, Spark, Kafka, HBase, and so on. If your use case requires it to be zone-redundant and handle AZ failures, you can use an HA cluster template and then use the Multi-AZ flag when you create the Data Hub clusters. For more information, refer to the AWS and Azure use case documents for limitations.

In a multiple availability zones (multi-AZ) setup, the Control Plane and the Data Hub service distribute each individual host across a host group across multiple AZs. For some services like Zookeeper, the service instances are distributed across multiple AZs regardless of the host group definitions.