Cassandra for System Design Interviews: When to Pick It, How It Shards, and Every Question You’ll Be Asked

A beginner-friendly guide to Apache Cassandra: what it is, why it scales, how it splits data across machines, and how to defend your choice when the interviewer starts asking “why?”

You’re twenty minutes into a system design interview. You’ve drawn the boxes for a chat app, and you say, confidently:

“I’ll store the messages in Cassandra.”

The interviewer leans in.

“Why Cassandra and not Postgres? How would you split the data across machines? What’s your partition key? What happens when a node dies in the middle of a write? Is Cassandra AP or CP?”

This is the moment many candidates lose points. Not because they picked the wrong database, but because they picked it by name, without the reasoning behind it.

This guide fixes that. By the end, you’ll be able to:

• Explain what Cassandra is in one sentence, and draw how it works.
• Say exactly when to choose it, and when not to.
• Explain sharding in Cassandra: partition keys, hashing, the token ring, and virtual nodes.
• Pick a good partition key, and fix a “hot” one.
• Explain replication, tunable consistency, and the famous R…