Apache Storm Interview Questions and Answers

Apache Storm is an open-source distributed computation system used to process unbounded streams of data in real-time.  Apache Storm is simple and it can be used with any programming language. It is used by many companies for real-time data analytics with fast data processing and fault tolerance. It is written in Clojure and Java Programming.

The key components of Apache Storm are listed below:

• Topology
• Streams
• Spouts
• Bolts

Apache Storm has master-slave architecture. In this architecture, the Nimbus running on a single node is called a master server. The superior running on each working node acts as a slave service. Apache Storm is a fault-tolerant and it consists of two nodes namely master node and worker node. The internal distributed messaging system is being used in Apache storm for building communication between supervisors and nimbus.

The essential components that play a key role in Apache Storm – Cluster Architecture are as follows:

• Nimbus
• Supervisor
• Executor
• Worker Process
• Task
• ZooKeeper Framework

The major differences between Apache Storm and Spark are as follows:

Apache Storm is a platform designed to work with real-time data. It runs on YARN and completely integrates with Hadoop Ecosystem. Storm runs different topologies composed of multiple components arranged in a Directed Acyclic Graph (DAG). The data between those components and each component consumes one or more data streams and can also emit one or more data streams. Spouts in Stor are used to bring data into a topology. Bolts are used to consume the emitted streams from Spouts and are also capable of writing data to external services. 

Storm Topology is defined as a graph of computation. Every node in the topology contains processing logic. All nodes in topology execute in parallel. Running a topology is done in a straightforward manner. Nodes are linked which indicates the data flow between them. 

 The significant difference between Apache Storm and Hadoop are as follows:

There are two different types of nodes available in Apache Storm, and they are as follows:

• Master Node: In Apache Storm, Nimbus is known as a Master Node. The main aim of the master node is to run storm topology. It is considered as a central component of Storm. It is used to gather the tasks to be executed and also used to analyze topology. It is also responsible for distributing data among all the worker nodes.

Worker Node: Supervisor is known as a worker node. The worker nodes follow the instructors given by a Nimbus. It also processes the tasks assigned by the nimbus and completes them in time. Both the Supervisor and Nimbus communicate with each other using an Internal Distributed Messaging System.

Apache Storm provides Guaranteeing Message Processing mechanism to ensure data guarantee processing even if the messages are lost or nodes die. 

In Apache Storm, the Spouts is prominently used as a source of streams in a topology. You can read tuples from an external source using Spouts and can emit them into a topology. The Spouts can either be reliable or unreliable. The Spouts have an in-built interface in it that can be used to run the specific logic of your application. In Apache Storm, the bolts are used to represent nodes in the topology. Bolts receive data from Spouts and emit it to one or more bolts. Bolt consists of the smallest processing unit. Moreover, the bolt is also used to execute multiple tasks in topology.