Getting Started with Streaming Analytics
Also available as:
PDF
loading table of contents...

Pick your Streaming Engine

Hortonworks supports a number of powerful Streaming Engines including:

  • Spark Streaming (Currently Supported in HDP)
  • Streaming Analytics Manager (SAM) using Apache Storm (Currently Supported in HDF)
  • Apache Kafka Streams (Will be supported in future HDP/HDF release)

Hortonworks provides you the flexibility to pick the streaming engine of your choice to build streaming analytics application.

If your organization has not standardized on a streaming engine and is looking for guidance on choosing an engine, use the below table to help guide your selection.

Requirement Streaming Engine to Use
You want to build streaming applications with as little code as possible and want to use ETL like Tooling with a drag and drop paradigm to build streaming apps Streaming Analytics Manager (SAM) with Storm
You want the ability to build an app that you can deploy in batch and/or streaming mode Spark Streaming
You plan to develop streaming applications using scala and java and want a clean easy to use API Spark Streaming
You want the ability to execute SQL against the stream

Spark Streaming

You want process event one at a time (no microbatching) Streaming Analytics Manager (SAM) with Storm

The below sections walk you through implementing the streaming analytics requirements with these different tools/engines.