Caching

🔁 Caching — Explained Simply

Caching is the process of storing copies of data in a temporary storage location (called a cache) so that future requests for that data can be served faster.

---

🚀 Why Caching?

Fetching data from the original source (like a database, API, or disk) is slow and expensive. Caching avoids this by keeping frequently accessed data in a fast-access layer (usually memory).

---

🧠 Real-life Example

Imagine you’re Googling “weather in Pune”:

• Without cache: It hits the weather server every time you search.

• With cache: If someone already searched this recently, the result is shown instantly from memory.

---

🧱 Types of Caching

Type	Where It Happens	Example
Browser Cache	On the client	Images, CSS, JS files
CDN Cache	On edge servers	Cloudflare serving static content
Application Cache	In app memory	Storing API responses
Database Cache	Beside DB	Redis/Memcached storing query results
Operating System Cache	At OS level	File system cache

---

🔄 How Caching Works (Basic Flow)

User Request
    ↓
Check Cache → Hit? → Serve from cache ✅
        ↓
     Miss? → Fetch from DB/API → Save in Cache → Return

---

🛠️ Tools Commonly Used

• Memory-based: Redis, Memcached

• Browser-level: LocalStorage, IndexedDB, HTTP headers (Cache-Control)

• CDNs: Cloudflare, Akamai, Fastly

---

⚠️ Cache Invalidation – The Hard Part

Biggest challenge is knowing when to update or delete stale cache. Strategies:

• Time-Based (TTL): Cache expires after x seconds

• Manual Invalidate: Clear when data changes (e.g., on DB write)

• Write-through Cache: Update cache immediately after writing to DB

• Stale-while-revalidate: Serve stale data while fetching updated data in the background

---

⚖️ Pros vs Cons

✅ Pros:

• Faster performance (lower latency)

• Reduces backend load

• Cost-effective at scale

❌ Cons:

• Hard to keep fresh (stale data issues)

• More system complexity

• Risk of serving outdated or inconsistent data

---

📦 When Should You Use Caching?

• Data is read-heavy

• Data doesn’t change frequently

• Performance is critical

• You want to reduce DB/API costs