Cache Invalidation — Beginner Friendly System Design Guide

Why Caching Creates Problems

Caching makes applications:

• faster
• cheaper
• scalable

But there is one big challenge:

What happens when original data changes?

If cache still shows old data:

Users see stale information

This problem is called:

Cache Invalidation

---

What Is Cache Invalidation?

Cache invalidation means:

Removing or updating old cached data when actual data changes.

---

Real Example — Blinkit Inventory

Suppose Blinkit has:

1 milk packet left

Inventory cached in Redis.

User A buys it.

Database updates:

Stock = 0

But cache still contains:

Stock = 1

Now User B also orders.

Problem:

• overselling inventory

---

Without Cache Invalidation

Database Updated
       │
       ▼
Old Cache Still Exists ❌
       │
       ▼
Users See Wrong Data

---

With Cache Invalidation

Database Updated
       │
       ▼
Invalidate Cache
       │
       ▼
Next Request Gets Fresh Data

---

Common Cache Invalidation Strategies

---

1. TTL (Time To Live)

Cache automatically expires after fixed time.

---

Example

Cache expires after 5 minutes

---

TTL Flow

Store Data In Cache
       │
       ▼
TTL Countdown Starts
       │
       ▼
Cache Auto Expires

---

Best For

• product lists
• news feeds
• homepage data

---

Problem

Users may still see:

• old data until expiry

---

2. Manual Invalidation

Application deletes cache immediately after data update.

---

Flow

Update Database
       │
       ▼
Delete Cache
       │
       ▼
Fresh Data Loaded Again

---

Best For

• inventory
• payments
• booking systems

---

Real Example — Swiggy

Restaurant changes:

OPEN → CLOSED

Cache must update quickly.

Otherwise:

• users may place invalid orders.

---

3. Cache-Aside Pattern

Most common industry approach.

Application:

• checks cache first
• queries database if cache miss happens

---

Read Flow

User Request
      │
      ▼
Check Cache
  │            │
Hit           Miss
 │              │
 ▼              ▼
Return       Query Database
Cache Data

---

Update Flow

Update Database
       │
       ▼
Invalidate Cache

---

Why Cache Invalidation Is Hard

Modern systems may have:

• browser cache
• CDN cache
• Redis cache

All can become stale.

---

Multi-Level Cache Flow

Browser Cache
      │
      ▼
CDN Cache
      │
      ▼
Redis Cache
      │
      ▼
Database

---

Real Example — YouTube Thumbnail Update

Creator changes thumbnail.

Need to invalidate:

• browser cache
• CDN cache
• backend cache

Otherwise users still see:

• old thumbnail

---

Common Beginner Mistakes

Mistake	Problem
Long cache expiry	Stale data
Cache everything	Wasted memory
Forget invalidation	Wrong user experience

---

Final Mental Model

Caching improves speed.
Invalidation keeps data correct.

Both are equally important.

---

One-Line Interview Definition

Cache invalidation is the process of removing or refreshing stale cached data when the original source of truth changes.