Course contentsShow
← System Design
Lesson 342 of 1,91910. NoSQL: Key-Value StoresFree lesson

No Secondary Indexes or Query Language

Understanding the limitation: you can only access data by its primary key, no SQL-like queries or secondary indexes.

No Secondary Indexes or Query Language

What you'll learn: Why key-value stores only let you retrieve data using the primary key, with no SQL-like queries or indexes.

The Core Limitation

In a key-value store, you can only access data by its exact key. Unlike relational databases where you can query by any indexed column or write complex SQL statements, key-value stores have no concept of secondary indexes, WHERE clauses, or JOINs. If you don't know the key, you can't efficiently find the value.

Why This Matters

Think of a key-value store like a massive dictionary. You can instantly look up a word (key) and get its definition (value). But what if you want to "find all words that start with 'pre-'" or "show me all definitions containing the word 'happiness'"? You'd have to flip through every single page—the dictionary has no index for that kind of search.

In practical terms:

  • ✅ Allowed: GET user:12345 retrieves a user by their ID
  • ❌ Not allowed: "Find all users where age > 30"
  • ❌ Not allowed: "Get all users in California"
  • ❌ Not allowed: "List users sorted by name"

Design Implications

This constraint forces you to design your keys carefully and often duplicate data under different keys if you need multiple access patterns. For example, if you need to look up users by both ID and email, you might store:

  • user:12345 → user data
  • email:jane@example.com → 12345 (points to the user ID)

This trades storage space for query simplicity and speed.

Key Takeaway: Key-value stores sacrifice query flexibility for extreme speed and simplicity—you must know the exact key to retrieve data, which shapes how you model and structure your application's data access patterns.