Working with Trident

Throughout the book, we have portrayed Storm as a solution to be used as a high-performant, real-time stream computational tool. But, in reality, all real-time use cases are not actually real-time; they are an extension to real-time and use an amalgamation of micro batching. Let's give you some examples.

Let's say we want to know the names of the top five performing stocks. This data should reflect the stock performance over the last 10 minutes. We also want to know which is the most liked picture on Facebook for each 5-minute period.

There are numerous scenarios that require small unit batching over real-time streaming data and computations around the same; thus, the need for the extension of Storm.

Trident, like its predecessor, Storm, was also incepted at the Twitter technology house. At a high level, it's an extension and abstraction over the top of the Storm framework, with some additional capability for batching, stateful processing, and querying of the streaming...