Cassandra Tutorial - Learn Cassandra from Scratch

Here’s a list of topics we’ll be covering in this blog:

• What is Apache Cassandra?
• History of Cassandra
• Features of Cassandra
• Applications of Apache Cassandra
• Comparison between NoSQL and Relational Database
• Cassandra Architecture
• Data Modeling in Cassandra
• What is a NoSQL Database?
• Why should you learn Cassandra?
• Recommended Audience
• Prerequisites

What is Apache Cassandra?

Apache Cassandra is an open-source, powerful, distributed NoSQL database that does not have a single point of failure and is extremely scalable and highly available. Cassandra was originally developed at Facebook and later open-sourced and is currently part of the Apache Software Foundation. Cassandra is basically a high-performance, high availability, and highly scalable distributed database that works well with structured, semi-structured, and unstructured data. For structured data we have the RDBMS, so a database like Cassandra is essentially used for collecting and handling unstructured data.

Apache Cassandra is a highly scalable, high-performance distributed database that can manage massive data volumes across a number of commodity computers, offering high availability with no single point of failure. It is a type of NoSQL database. Let us first understand how the NoSQL database work. A distributed database management system called Cassandra is made to manage a large amount of structured data over a wide range of commodity computers.

Cassandra’s distributed architecture manages a massive amount of data. Since there is no single point of failure and data is spread across multiple machines, high availability is achieved.

History of Cassandra

Cassandra was first created at Facebook by Prashant Malik and Avinash Lakshman, one of the inventors of Amazon’s Dynamo, to support the Facebook inbox search feature. In July 2008, Facebook published Apache Cassandra as an open-source or publicly accessible project on Google Code. It was accepted as an Apache Incubator project in March 2009. The project advanced to a top-level project on February 17, 2010.

With historical references to a curse on an oracle, Facebook’s database was given the name Cassandra by its creators in honor of the Trojan fabled prophet. Facebook originally created Cassandra for inbox search.

• Facebook released the code in July 2008.
• Cassandra was recognized by the Apache incubator in March 2009.
• Cassandra has been an Apache top-level project since February 2010.
• 3. 2.1 is the most recent release of Apache Cassandra.

Features of Cassandra

Java Management Extensions can be used to administer and track Cassandra, a Java-based system (JMX). To operate a Cassandra cluster, for example, one can utilize the JMX-compliant nodetool software (adding nodes to a ring, draining nodes, decommissioning nodes, and so on).

Additionally, Nodetool provides several commands that can retrieve Cassandra metrics for things like disc consumption, latency, compaction, trash collection, and more.

Here in this section of the Cassandra tutorial, we will discuss some of the top features of Cassandra

• Scalability: Cassandra is highly scalable meaning you can have additional hardware for accommodating more customers and data
• Better Architecture: Cassandra does not have a single point of failure and it has an always-on architecture
• Performance: It has a fast linear performance which means you can increase the throughput by increasing the number of nodes in the cluster
• Storage: It has a highly flexible data storage meaning all formats of data can be stored including structured, semi-structured, and unstructured
• Distribution: It allows for easy data distribution by providing the flexibility to distribute data by replicating it across multiple data centers
• ACID Properties: Cassandra supports the ACID compliance which stands for Atomicity, Consistency, Isolation, Durability
• Efficiency: It performs blazing fast writes without sacrificing the read efficiency.
• Fault Resistance: When all nodes are treated equally, it doesn’t matter too much if one goes down. You may essentially increase the number of nodes so that you will never have a complete “lights out” situation.
• Consistency: On top of standard JVM performance tuning, Cassandra offers a lot more flexibility.

Applications of Apache Cassandra

Apache Cassandra is one of the most widely used NoSQL databases. Here we list some of the top applications of Cassandra.

• It is extensively used for monitoring and tracking applications.
• It is used in web analytics which are heavy write systems.
• It is deployed for social media analysis for providing suggestions to customers.
• It is used in retail applications for product catalog lookups and inputs.
• It is extensively used as the database for mobile messaging services.

Comparison between NoSQL and Relational Database

Let us first understand the difference between a NoSQL database and a relational database through this table:

Apache Cassandra is an open-source, powerful, distributed NoSQL database that does not have a single point of failure and is extremely scalable and highly available. Cassandra was originally developed at Facebook and later open-sourced and is currently part of the Apache Software Foundation.

Cassandra Architecture

Cassandra is designed to handle Big Data workloads. It is capable of doing that across different nodes without failure at any point. Cassandra has a peer-to-peer system that is distributed across its multiple nodes. The data is distributed in a cluster among these nodes. Data dissemination to nodes is made transparent by the Cassandra design. This means that you can use the data to establish where your data is located within the cluster. There are no masters or slaves, therefore any node can accept any request. A node will return the data if it has it. If not, the request will be sent to the node that contains the data.

To obtain the necessary amount of redundancy, you can define the number of data replicas. For instance, you might wish to provide a replication factor of 4 or 5 if the data is extremely important.

• All nodes present within a cluster play the same role. All the nodes are interconnected to each other and yet independent.
• All the nodes are capable of accepting read and write requests. This isn’t dependent on where the data is located in the cluster.
• Read/write requests can be served to other nodes if a particular node goes down.
• Cassandra was created without a master or slave node.
• Its nodes are logically distributed in a way that resembles a ring, making it have a ring-type architecture.
• All the nodes receive data distribution automatically.
• Data is copied between the nodes for redundancy, just like HDFS.
• Data is stored in memory and is only sometimes written to the disc.
• The data is distributed among cluster nodes using the hash values of the keys

Data Modeling in Cassandra

Cassandra data model differs greatly from the typical RDBMS data model. A general overview of Cassandra’s data storage is given in this chapter.

• Clustering: The Cassandra database is spread over a number of interconnected machines. The Cluster is the name given to the outermost container.
• KeySpace: Keyspace is the outermost layer for data in Cassandra.
• Column Family: An ordered group of rows is included within a column family. Each row is composed of a set of columns in a particular sequence. The elements that set a column family apart from a relational database table are listed in the following table.
• Column: Cassandra’s fundamental data structure is a column, which has three values: a key, which is the column name, a value, and a time stamp. The structure of a column is seen below.
• SuperColumn: Being a special column, a super column is also a key-value pair. A super column, however, keeps a map of the sub-columns.

What is a NoSQL database?

A NoSQL or Not Only SQL is a set of databases that provide a way to store and retrieve data that is not in the standard tabular format followed by relational databases. The NoSQL databases of which Cassandra is a very popular database share some common features and attributes. The NoSQL databases do not have any schema, they support easy replication of data, they have a simple API, they do not exhibit the ACID properties but are eventually consistent, and finally last but not least, they can handle huge volumes of data.

Some of the properties of a NoSQL database include:

• It has a simple design.
• It is scalable horizontally.
• It has finer control over availability.

Why should you learn Cassandra?

Cassandra is a top NoSQL database and it is finding more and more users with each passing day. Since we are living in a world of Big Data, Cassandra is finding huge acceptance since it was built for Big Data. Also, a lot of the organizations are moving from the traditional relational database systems to NoSQL databases and thus, Cassandra is their natural choice.

All this means that the job market for Cassandra is just heating up and the salaries for Cassandra professionals are among the best in the Big Data domain. All these are compelling reasons for you to learn Cassandra and excel in your career.

Let’s look at some of the major points why Cassandra is such a widely used NoSQL database.

• It is a high-performance and high availability database.
• It is extremely fault-tolerant, scalable, and consistent.
• It is high-speed, thanks to it being a column-oriented database.
• Its architecture is based on Google’s Bigtable & Amazon’s Dynamo.
• It can manage extremely large data sets.

Recommended Audience

This Cassandra tutorial can be beneficial to anybody who wants to learn NoSQL databases. Software developers, database administrators, architects, managers can take this Cassandra tutorial as a first step to learn Cassandra and excel in their careers.

Prerequisites

There are no prerequisites to learning Cassandra from this Cassandra tutorial. If you have a basic knowledge of databases, then it is good.