Replication and Sharding

---

1. Introduction

Modern applications require databases that are highly available, fault tolerant, and scalable.
Replication and sharding are two important techniques used in NoSQL databases to achieve these goals.

---

2. Replication

Definition

Replication is the process of copying the same data to multiple servers (nodes).

---

Purpose of Replication

• Improves data availability

• Provides fault tolerance

• Increases read performance

• Prevents data loss

---

Types of Replication

1. Master–Slave Replication

• One master node handles writes

• Slave nodes replicate data from master

• Reads can be performed from slaves

2. Peer-to-Peer Replication

• All nodes are equal

• Data is replicated across nodes

• No single point of failure

---

Advantages of Replication

• High availability

• Data backup

• Better system reliability

---

Challenges of Replication

• Data consistency issues

• Replication delay

• Conflict resolution

---

3. Sharding

Definition

Sharding is the process of dividing large datasets into smaller parts (shards) and storing them on different servers.

Each shard contains only a portion of the data.

---

Purpose of Sharding

• Improves scalability

• Distributes data and workload

• Handles large volumes of data

---

Sharding Methods

• Range-based sharding

• Hash-based sharding

• Directory-based sharding

---

Advantages of Sharding

• Horizontal scalability

• Better write performance

• Efficient data distribution

---

Challenges of Sharding

• Complex data management

• Query routing complexity

• Data rebalancing

---

4. Difference Between Replication and Sharding

Aspect	Replication	Sharding
Purpose	Availability	Scalability
Data Stored	Same data on multiple nodes	Different data on each node
Failure Handling	Yes	Limited
Performance	Improves reads	Improves writes

---

5. Replication and Sharding in NoSQL

• NoSQL databases support automatic replication and sharding

• Ensures high availability and scalability

• Used together in large distributed systems

Examples:

• MongoDB – Replica sets & sharding

• Cassandra – Built-in replication and partitioning