Key-Value Database

Definition: A key-value database is a type of NoSQL database that stores information as pairs: a unique key points directly to a value.

It is especially useful when you already know the key and need to retrieve data very quickly, rather than performing complex queries or joins. (AWS Builder Center)

Simple example

Think of a dictionary:

"username:123" → "Reyhaneh"
"session:abc"  → "user_id=123"
"cart:123"     → ["laptop", "mouse"]

You give the database the key:

GET "session:abc"

and it gives you the value:

"user_id=123"

The database doesn’t need to search through lots of rows to find it—the key directly identifies the value.

3 examples

1. User sessions

   session:8f92 → {user_id: 42, logged_in: true}

   When the user makes a request, you can quickly retrieve their session.

2. Caching

   product:123 → {name: "Laptop", price: 1200}

   Instead of repeatedly querying PostgreSQL, you can temporarily keep frequently used data in a key-value store. (Couchbase)

3. Gaming leaderboard

   player:987 → score: 15420

   The application can quickly retrieve or update a player’s score. (Dragonfly)

Common examples

• Redis — very fast, commonly used for caching, sessions, and real-time data. (Redis)

• Amazon DynamoDB — a managed database supporting key-value and document models. (Amazon Web Services, Inc.)

• Memcached — commonly used for simple in-memory caching. (techtarget.com)

Easy way to remember:

Relational DB: “Find all users whose age is greater than 20.”
Document DB: “Give me this user’s document.”
Key-Value DB: “I know the key → just give me its value.”

---

Database My-Journey-In-Codeless