What is Apache Cassandra and what problems does it solve?

Expected Interview Answer

Step-by-Step Explanation

1. Step 1

   Identify the scaling need

   You have write-heavy workloads and data volumes too large for a single relational server to handle.

2. Step 2

   Choose a distributed model

   Cassandra partitions data across a ring of equal nodes using consistent hashing on a partition key.

3. Step 3

   Replicate for safety

   Each partition is copied to a configurable number of nodes (the replication factor) across racks and data centers.

4. Step 4

   Tune consistency

   Pick per-query consistency levels (ONE, QUORUM, ALL) to balance latency against how many replicas must agree.

5. Step 5

   Scale out

   Add commodity nodes to grow capacity and throughput linearly, with no downtime and no master to overload.

Code Example

CREATE KEYSPACE store
  WITH replication = {
    'class': 'NetworkTopologyStrategy',
    'datacenter1': 3
  };

CREATE TABLE store.users (
  user_id UUID PRIMARY KEY,
  name TEXT,
  email TEXT
);

Follow-up Questions

• What is tunable consistency in Cassandra?
• How does Cassandra partition and replicate data?
• What is the gossip protocol?
• When should you NOT use Cassandra?
• What is a partition key versus a clustering key?