Storm is a distributed, scalable, fault-tolerant real-time computing system open sourced by Twitter. Just as Hadoop greatly simplifies parallel batch data processing, storm defines a batch of real-time computing primitives that greatly simplify parallel real-time data processing. In terms of overall architecture, storm is very similar to Hadoop and solves the problem of poor real-time performance of Hadoop [8, 9]. Therefore, it is also referred to as a “real-time Hadoop” system. It can be said that Storm's real-time processing is like Hadoop's batch processing. Table 2 shows the correspondence and similarities between Hadoop and Storm from three perspectives: system roles, application names, and component interfaces.