Scribd
Upload
46 views4 pages

Understanding Sharding in NoSQL Databases

Sharding is a technique for distributing data across multiple servers to enhance performance and scalability, particularly in NoSQL databases. It involves dividing data into smaller parts called shards, each managed by a separate server, and uses a shard key to determine data allocation. While sharding improves throughput and avoids single points of failure, it also presents challenges such as complex setup and the need for careful shard key selection.

Uploaded by

Akanksha Rajput
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
46 views4 pages

Understanding Sharding in NoSQL Databases

Sharding is a technique for distributing data across multiple servers to enhance performance and scalability, particularly in NoSQL databases. It involves dividing data into smaller parts called shards, each managed by a separate server, and uses a shard key to determine data allocation. While sharding improves throughput and avoids single points of failure, it also presents challenges such as complex setup and the need for careful shard key selection.

Uploaded by

Akanksha Rajput
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd

1. What is Sharding?

Sharding is a method of distributing data across multiple servers (or


nodes) to improve performance, scalability, and storage capacity in a
database system — especially in NoSQL databases like MongoDB,
Cassandra, and HBase.

It’s like dividing a big table into smaller parts (shards) so that each
server handles only a portion of the data.

2. Why Sharding is Needed

As data grows, a single server may face problems:

 Storage limits (too much data)


 Slow queries (too many requests)
 High load (CPU and memory usage)

Sharding solves this by splitting data horizontally — each shard stores


a subset of rows (not columns).
3. How Sharding Works

1. Shard Key – A field (or combination of fields) used to decide


which shard a document or record belongs to.
Example: In a MongoDB “students” collection, student_id
could be the shard key.
2. Shards – Each shard is a separate database server (or cluster)
containing part of the data.
3. Config Server / Metadata – Keeps track of which shard holds
which data ranges.
4. Router (Query Router) – Routes queries to the correct shard(s)
based on the shard key.

4. Example

Suppose we have a NoSQL collection:

student_id name branch city


1 Raj CSE Gwalior
2 Neha ECE Delhi
3 Aman ME Pune

If we shard by student_id:

 Shard 1: student_id 1–1000


 Shard 2: student_id 1001–2000
 Shard 3: student_id 2001–3000

When a query asks for student_id = 1200, the router sends it


directly to Shard 2 only.

5. Advantages of Sharding

Scales horizontally (add more servers easily)


Increases read/write throughput
Distributes data evenly across multiple nodes
Avoids single point of failure (in distributed setups)
6. Challenges / Disadvantages

Complex setup and maintenance


Requires careful choice of shard key (bad key → uneven load)
Data migration between shards can be costly
Some queries (that don’t use shard key) must scan all shards

7. NoSQL Databases That Use Sharding

 MongoDB – Built-in automatic sharding


 Cassandra – Uses consistent hashing (automatic sharding)
 HBase – Region-based sharding
 Couchbase – Uses vBuckets for sharding

🧠 Summary

Concept Description
Splitting large data horizontally into smaller chunks stored
Sharding
on different servers
Shard
Attribute used to decide which shard a record belongs to
Key
Goal Improve scalability, performance, and data distribution

You might also like