Streaming Analytics Data Hub cluster definitons
There are four Streaming Analytics cluster definitions available to deploy Streaming Analytics in CDP Public Cloud. You can choose from Light and Heavy duty options, and you can further select the cluster definitions depending on your cloud provider.
- Streaming Analytics Light Duty for AWS
- Streaming Analytics Light Duty for Azure
- Streaming Analytics Heavy Duty for AWS
- Streaming Analytics Heavy Duty for Azure
Streaming Analytics offers real-time stream processing and stream analytics with low-latency and high scaling capabilities powered by Apache Flink.
Streaming Analytics templates include Apache Flink that works out of the box in stateless or heavy state environments. Beside Flink, the template includes its supporting services namely YARN, Zookeeper and HDFS. The Heavy Duty template comes preconfigured with RocksDB as state backend, while Light Duty clusters use the default Heap state backend. You can create your streaming application by choosing between Kafka, Kudu, and HBase as datastream connectors.
In CDP Public Cloud, you can connect to the Streams Messaging cluster to create an end-to-end streaming environment including Streams Messaging Manager and Schema Registry beside Kafka. You can also use the Real-time Data Mart cluster definition to have Kudu in your environment. To further enhance the streaming analytics experience, the SQL Client is also available for creating analytical summaries and data queries. You can use the Flink SQL Client when Kafka and Kudu are chosen as connectors.