A Comparison of Big Data Frameworks on a Layered Dataflow Model

TL;DR

This paper compares various Big Data frameworks using a layered Dataflow model, demonstrating their expressiveness and providing a unified understanding of their capabilities and abstractions.

Contribution

It introduces a layered Dataflow model that captures the expressiveness of existing frameworks like Spark, Flink, and Storm, clarifying their relationships and capabilities.

Findings

The model is as general as existing frameworks.

It unifies understanding of batch and streaming tools.

Tools fit into the layered Dataflow structure.

Abstract

In the world of Big Data analytics, there is a series of tools aiming at simplifying programming applications to be executed on clusters. Although each tool claims to provide better programming, data and execution models, for which only informal (and often confusing) semantics is generally provided, all share a common underlying model, namely, the Dataflow model. The Dataflow model we propose shows how various tools share the same expressiveness at different levels of abstraction. The contribution of this work is twofold: first, we show that the proposed model is (at least) as general as existing batch and streaming frameworks (e.g., Spark, Flink, Storm), thus making it easier to understand high-level data-processing applications written in such frameworks. Second, we provide a layered model that can represent tools and applications following the Dataflow paradigm and we show how the analyzed tools fit in each level.