What is Scalability in System Design?

System Design Refresher

1. Instagram System Design

2. System Design: The Distributed Task Scheduler

3. System Design of Reddit

4. 10 System design trade-offs

5. Rate Limiting Algorithms Explained with Code

6. Low-Level Design (LLD) for Better Software Solutions

7. Stateful vs Stateless Architecture

8. Best Practices for Developing Microservices

9. 10 Problems of Distributed Systems

10. 20 System Design Concepts Every Developer Should Know - Part - I

11. How Shopify handles 16,000 Request per second

12. Software Architecture Pattern - Layered Architecture

13. How Enterprise Applications Exchange Data Using Messaging

14. Microservices Design Pattern - Event Sourcing Pattern

15. Improve API Performance 🚀

16. Distributed System Learning Roadmap

17. 15 Scaling Techniques in Distributed Systems

---

Scalability

15 Scaling Techniques in Distributed Systems

Scalability is the ability of a system to handle increased load (users, data, or requests) without sacrificing performance or reliability.

Two main approaches:

1. Horizontal Scaling (Scale Out)

   • Add more machines/servers.

   • Example: Adding more web servers behind a load balancer when traffic spikes.

2. Vertical Scaling (Scale Up)

   • Add more power (CPU, RAM, storage) to a single machine.

   • Example: Upgrading your database server to a larger instance.

👉 Horizontal scaling is often preferred for distributed systems because it’s cost-effective and avoids single points of failure.

---

Sharding

Deep Dive in Data Partitioning & Sharding

Sharding is splitting a large database into smaller, more manageable chunks called shards. Each shard stores only a portion of the data.

• How it works:

  • Users 1–1M → Shard A

  • Users 1M–2M → Shard B

  • Users 2M–3M → Shard C

• Why it matters:

  • Avoids overwhelming a single database.

  • Improves read/write performance.

  • Enables scaling out databases horizontally.

• Real-world example:

  • Twitter shards user data so that one database isn’t responsible for every tweet ever posted.

---

Replication

Replication means creating multiple copies of the same data across different servers.

• Models:

  1. Master-Slave:

     • Master handles writes.

     • Slaves handle reads.

  2. Master-Master:

     • Multiple masters handle both reads/writes, syncing changes between them.

• Benefits:

  • High availability (failover if one DB goes down).

  • Improved read performance (scale reads across replicas).

  • Data durability (extra copies).

• Example:

  • MySQL replication powers many e-commerce platforms to ensure reads are fast and failover is automatic.

---

Partitioning

Partitioning is similar to sharding but usually refers to splitting data within a database (logical or physical divisions).

• Types of Partitioning:

  1. Horizontal Partitioning (Sharding) – Rows are divided across different databases/tables.

  2. Vertical Partitioning – Columns are split into different databases/tables (e.g., separating user profile info from login credentials).

• Benefits:

  • Reduces table size → faster queries.

  • Isolates workloads (analytics vs transactions).

---

Caching

Caching — The Only Guide You’ll Ever Need

Caching stores frequently accessed data in a faster, temporary storage layer (often memory). This reduces database load and improves response times.

• Strategies:

  1. Cache-aside (Lazy Loading): App checks cache → if miss, fetch from DB → update cache.

  2. Write-through: Writes go to both cache + DB.

  3. CDN (Content Delivery Network): Geographically distributed caches for static files (images, videos, CSS).

• Why caching matters:

  • Faster responses.

  • Lower infrastructure costs (fewer DB hits).

  • Better user experience during high traffic.

• Example:

  • Netflix caches video metadata at edge servers so users can browse instantly.