Cassandra Introduction

Apache Cassandra is a free, open-source database system built to handle massive amounts of data across many computers at the same time. Big companies like Netflix, Instagram, and Apple use it to keep their services running even when millions of people are active at once.

What Problem Does Cassandra Solve?

Imagine you run a popular food delivery app. On a regular Tuesday, 10,000 orders come in per hour. On New Year's Eve, that number jumps to 500,000 orders per hour. A traditional database would crash or slow down badly. Cassandra handles this by spreading the load across many machines, so no single machine carries all the weight.

The Restaurant Analogy

Think of a traditional database as one very skilled chef cooking all orders alone. When the restaurant gets busy, orders pile up and customers wait. Cassandra works like a kitchen with 50 chefs working side by side. Each chef handles a portion of the orders, so service stays fast no matter how full the restaurant gets.

Traditional Database (One Chef):
  [All Requests] ──▶ [Single Server] ──▶ [Slow / Crash]

Cassandra (Many Chefs):
  [All Requests]
       │
       ├──▶ [Node 1] ──▶ fast
       ├──▶ [Node 2] ──▶ fast
       ├──▶ [Node 3] ──▶ fast
       └──▶ [Node 4] ──▶ fast

Key Characteristics of Cassandra

1. Distributed

Cassandra stores data across multiple computers called nodes. These nodes work together as a team. You never rely on just one machine, so a hardware failure does not bring down your entire system.

2. No Single Point of Failure

Every node in Cassandra holds a copy of the data. If one node goes offline, other nodes immediately serve the same data. Your application keeps running without interruption.

3. Linear Scalability

Adding more nodes increases Cassandra's capacity in a predictable way. Double the nodes and you roughly double the throughput. This is called linear scalability, and it is one of Cassandra's most powerful features.

4. High Write Speed

Cassandra excels at writing data very fast. It writes to memory first and then flushes data to disk in the background. This design makes it ideal for logging, tracking, and real-time event recording.

Where Did Cassandra Come From?

Facebook engineers created Cassandra in 2007 to power the inbox search feature. The project became open source in 2008 and joined the Apache Software Foundation in 2010. Today, the Apache Cassandra project is maintained by a global community of developers.

Who Uses Cassandra?

Company          Use Case
──────────────────────────────────────────
Netflix          Storing streaming data and user watch history
Apple            Handling billions of daily requests across services
Instagram        Managing user activity feeds
Uber             Recording real-time trip and location data
Twitter          Storing tweets and timelines

Cassandra vs a Spreadsheet

A spreadsheet stores data in neat rows and columns on one file. Cassandra stores data in tables too, but those tables live across many machines simultaneously. You can also store millions of rows without any performance penalty — something a spreadsheet could never do.

The CAP Theorem and Cassandra

The CAP theorem says a distributed database can only guarantee two of three things: Consistency, Availability, and Partition Tolerance. Cassandra prioritizes Availability and Partition Tolerance. This means Cassandra stays up and accessible even when network problems occur between nodes, though in rare edge cases you might read slightly outdated data. You can tune this behavior based on your application's needs.

What Cassandra Is Not

Cassandra does not support complex joins between tables the way a traditional relational database does. It does not run transactions that span multiple rows by default. It works best when you design your data model around the specific queries your application needs to run.

Summary

Cassandra solves the problem of storing and retrieving huge volumes of data reliably and at high speed. It distributes data across many nodes, keeps copies on multiple machines, and scales out by simply adding more servers. The rest of this course teaches you how to set it up, model your data correctly, and use it effectively in real applications.