Understand Analytics Processors
Analytics processors contain the logic to process and analyze the data gathered from your sensors and devices.
Use analytics processors to define how to process the device data and events. An application can have multiple analytics processors.
When you create an analytics processor you define the properties to generate the source template that you’ll use to implement the logic. When you generate the template you also generate the underlying infrastructure to run the analytics processor’s code. Therefore, before generating the source template, you must ensure that you’ve properly defined the links associated to that analytics processor.
From your analytics processors you can read and write to the links associated to it, you can use the classes and methods defined in the Java Spark APIs, and you can also use the classes and methods from the libraries that you imported to your IoT application.
The type of analytics processor to use depends on the nature of your data:
-
Streaming Analytics Processors
Use streaming analytics processors to analyze high volumes of rapidly changing data streamed from your sensors and devices. Streaming analytics processors are continuously running from the moment you deploy them, waiting for your devices to stream data.
-
Batch Analytics Processors
Use batch analytics processors to analyze large volumes of batch data. Typically, batch analytics processors process data stored in the Oracle NoSQL or other data sources. Unlike streaming analytics processors they are not continuously running, they only run when you call them using the REST API they expose.