Apache Cassandra Interview Questions

Apache Cassandra is a distributed NoSQL database designed to store and manage large volumes of data with high availability, scalability and fault tolerance. It covers:

• Fundamental concepts, architecture and core components of Cassandra.
• Data modeling, partitioning, replication and consistency mechanisms.
• Storage, read/write operations and advanced Cassandra features.
• Performance optimization, fault tolerance and distributed data management.

1. What is Apache Cassandra?

Apache Cassandra is an open-source, distributed NoSQL database designed to handle large volumes of structured and semi-structured data across multiple servers. It provides high availability, fault tolerance and horizontal scalability without a single point of failure. Cassandra follows a peer-to-peer architecture where every node in the cluster has the same responsibility, allowing applications to remain available even if one or more nodes fail.

2. What are the Key Features of Apache Cassandra?

Apache Cassandra provides several features that make it suitable for handling large-scale distributed applications.

• Distributed peer-to-peer architecture.
• High availability with no single point of failure.
• Horizontal scalability by adding new nodes.
• Automatic data replication across multiple nodes.

3. How Does Apache Cassandra Differ from Relational Databases?

4. Explain the Architecture of Apache Cassandra.

Apache Cassandra follows a peer-to-peer distributed architecture, where every node in the cluster has the same responsibility. The architecture consists of the following components:

• Node: Stores data and handles read and write requests.
• Cluster: A collection of nodes working together.
• Data Center: A group of nodes located in the same region.
• Commit Log: Records write operations before storing them in memory.
• Memtable: Stores data temporarily in memory before writing it to disk.
• SSTable: Stores immutable data on disk.
• Gossip Protocol: Enables nodes to exchange cluster information.
• Replication: Copies data across multiple nodes for high availability.

5. What is a Node in Cassandra?

A node is an individual server or machine that stores data and participates in a Cassandra cluster. Each node stores a portion of the data, processes read and write requests and communicates with other nodes in the cluster. Since Cassandra follows a peer-to-peer architecture, every node performs the same role and contributes equally to the cluster.

6. What is a Cluster in Cassandra?

A cluster is a collection of one or more Cassandra nodes that work together to store and manage data. All nodes in a cluster communicate with each other and share information using the peer-to-peer architecture. As storage or traffic requirements increase, additional nodes can be added to the cluster without affecting application availability.

7. What is a Data Center in Cassandra?

A data center is a logical group of Cassandra nodes, usually located in the same geographical region. Cassandra supports multiple data centers to improve fault tolerance, disaster recovery and low-latency access for users in different locations. Data can be replicated across multiple data centers based on the configured replication strategy.

8. What is a Keyspace in Apache Cassandra?

A keyspace is the top-level container in Apache Cassandra that stores tables. It is similar to a database in relational database management systems. A keyspace also defines how data is replicated across the cluster by specifying the replication strategy and replication factor.

Example

CREATE KEYSPACE company
WITH replication = {
    'class': 'SimpleStrategy',
    'replication_factor': 3
};

The above query creates a keyspace named company using the SimpleStrategy replication strategy with a replication factor of 3, meaning each partition is replicated across three nodes.

9. What is a Table in Apache Cassandra?

A table is a collection of related rows stored within a keyspace. Each table consists of columns and a primary key that uniquely identifies each row. Cassandra tables are designed based on application query patterns rather than following normalization rules.

Example

USE company;

CREATE TABLE employees (
    emp_id INT PRIMARY KEY,
    name TEXT,
    department TEXT,
    salary INT
);

The above query creates an employees table with four columns. The emp_id column is defined as the primary key and uniquely identifies each employee.

10. What is a Primary Key in Cassandra?

A primary key uniquely identifies each row in a Cassandra table. It consists of a partition key and, optionally, one or more clustering columns. The partition key determines how data is distributed across the cluster, while clustering columns determine the order of rows within a partition.

Example

CREATE TABLE employees (
    department TEXT,
    emp_id INT,
    name TEXT,
    salary INT,
    PRIMARY KEY (department, emp_id)
);

In the above table, department is the partition key and emp_id is the clustering column. Cassandra distributes data based on the partition key and stores rows within each partition in the order defined by the clustering column.

11. What is Cassandra Query Language (CQL)?

Cassandra Query Language (CQL) is the query language used to interact with Apache Cassandra. It provides an SQL-like syntax for creating keyspaces and tables, as well as inserting, updating, deleting and retrieving data. Although its syntax is similar to SQL, CQL is designed to work with Cassandra's distributed architecture.

Example

SELECT * FROM employees;

The above query retrieves all records from the employees table.

12. What are Partition Keys and Clustering Keys in Cassandra?

A partition key determines how data is distributed across the nodes in a Cassandra cluster, while a clustering key determines the order in which rows are stored within a partition. Together, they form the primary key and help optimize data distribution and query performance.

Example

CREATE TABLE employees (
    department TEXT,
    emp_id INT,
    name TEXT,
    salary INT,
    PRIMARY KEY (department, emp_id)
);

In the above table, department is the partition key and emp_id is the clustering key.

13. How Does Data Partitioning Work in Apache Cassandra?

Data partitioning is the process of distributing data across multiple nodes in a Cassandra cluster. Cassandra applies a hash function to the partition key to determine the node where the data will be stored. This ensures balanced data distribution, improves scalability and prevents a single node from becoming overloaded.

14. What is the Replication Factor in Cassandra?

The replication factor defines the number of copies of each partition stored across different nodes in a Cassandra cluster. A higher replication factor increases data availability and fault tolerance by ensuring that multiple copies of the same data are available even if one or more nodes fail.

Example:

CREATE KEYSPACE company
WITH replication = {
    'class':'SimpleStrategy',
    'replication_factor':3
};

In the above example, each partition is replicated across three different nodes.

15. What are Consistency Levels in Cassandra?

Consistency levels determine how many replica nodes must acknowledge a read or write operation before it is considered successful. Cassandra provides tunable consistency, allowing applications to choose between higher consistency and better performance based on their requirements.

Some commonly used consistency levels include:

• ONE: One replica must respond.
• QUORUM: A majority of replicas must respond.
• ALL: All replicas must respond before the operation succeeds.

16. Explain the Gossip Protocol in Apache Cassandra.

The Gossip Protocol is a peer-to-peer communication mechanism used by Cassandra nodes to exchange information about the cluster. It helps nodes discover each other, detect node failures and maintain the latest cluster state without relying on a central coordinator.

17. What are Hinted Handoffs in Cassandra?

Hinted Handoff is a mechanism that temporarily stores write operations when a replica node is unavailable. Once the unavailable node becomes active again, Cassandra automatically delivers the missed updates, helping maintain data consistency without interrupting write operations.

18. What is Read Repair in Cassandra?

Read Repair is a mechanism that helps maintain data consistency during read operations. If Cassandra detects different versions of the same data on replica nodes, it automatically updates the outdated replicas with the latest data to ensure consistency across the cluster.

19. What is the Difference Between Memtable and SSTable?

Memtable and SSTable are two storage components used by Cassandra during write operations.

20. What is Compaction in Cassandra and Why is it Required?

Compaction is the process of merging multiple SSTables into a single SSTable. It removes obsolete data, combines duplicate records and reduces the number of disk reads required during queries. Regular compaction improves storage efficiency and enhances read performance.

21. Explain the Write Path in Apache Cassandra.

The write path describes how Cassandra processes a write request before storing it on disk.

• The write request is first recorded in the Commit Log to ensure durability.
• The data is then written to the Memtable in memory.
• Once the Memtable reaches a threshold, it is flushed to an SSTable on disk.
• This process provides high write performance and prevents data loss.

22. Explain the Read Path in Apache Cassandra.

The read path describes how Cassandra retrieves data from storage.

• Cassandra first checks the Memtable for the requested data.
• If the data is not found, it searches the SSTables on disk.
• Bloom Filters help identify whether an SSTable may contain the required data.
• The latest data is returned to the client after checking the available replicas.

23. What are Tombstones in Cassandra?

A tombstone is a marker that represents deleted data.

• Created when data is deleted or expires using TTL.
• Prevents deleted data from reappearing during replication.
• Removed permanently during the compaction process.
• Helps maintain consistency across replica nodes.

24. What are Secondary Indexes in Cassandra?

Secondary indexes allow queries on columns that are not part of the primary key.

• Improve query flexibility.
• Useful for low-cardinality columns.
• Avoid using them on large datasets as they can reduce performance.

Example:

CREATE INDEX idx_department
ON employees(department);

25. What are Materialized Views in Cassandra?

Materialized Views store the same data as a base table using a different primary key.

• Support multiple query patterns.
• Automatically stay synchronized with the base table.
• Reduce the need to duplicate data manually.

Example:

CREATE MATERIALIZED VIEW employees_by_department AS
SELECT *
FROM employees
WHERE department IS NOT NULL
AND emp_id IS NOT NULL
PRIMARY KEY (department, emp_id);

26. What is Time to Live (TTL) in Cassandra?

Time to Live (TTL) specifies how long data should remain in a Cassandra table. After the specified time expires, Cassandra automatically marks the data for deletion.

• Automatically expires data after a specified duration.
• Useful for temporary or session data.
• Expired data is removed during compaction.

Example:

INSERT INTO employees (emp_id, name)
VALUES (101, 'John')
USING TTL 3600;

27. What are Lightweight Transactions (LWT) in Cassandra?

Lightweight Transactions (LWT) provide conditional updates using the Paxos protocol. They ensure that a write operation succeeds only if a specified condition is satisfied.

• Provides conditional INSERT, UPDATE and DELETE operations.
• Ensures strong consistency for critical operations.
• Useful for preventing duplicate records.

Example:

INSERT INTO employees (emp_id, name)
VALUES (101, 'John')
IF NOT EXISTS;

28. What are Batch Statements in Cassandra?

Batch statements allow multiple INSERT, UPDATE or DELETE operations to be executed as a single request.

• Reduces network communication between the client and server.
• Groups multiple operations into one request.
• Best suited for related write operations.

Example:

BEGIN BATCH

INSERT INTO employees (emp_id, name)
VALUES (101, 'John');

INSERT INTO employees (emp_id, name)
VALUES (102, 'Emily');

APPLY BATCH;

29. How Does Cassandra Handle Node Failures?

Apache Cassandra automatically handles node failures to ensure continuous availability of data.

• Uses data replication across multiple nodes.
• Detects failed nodes using the Gossip Protocol.
• Synchronizes missed updates using Hinted Handoff and Read Repair.
• Continues serving requests from available replica nodes.

30. How Can You Optimize Query Performance in Cassandra?

Query performance can be improved by following Cassandra's data modeling best practices.

• Choose an appropriate partition key.
• Avoid large partitions.
• Create indexes only when necessary.
• Perform regular compaction and repair operations.
• Design tables based on query patterns.