Course contentsShow
← System Design
Lesson 522 of 1,91915. Consistency ModelsFree lesson

What is Strong Consistency?

Definition of strong consistency: all reads see the most recent write, as if there's a single copy of data.

What is Strong Consistency?

You'll learn: What strong consistency means and why it makes distributed systems behave like a single, always up-to-date copy of data.

The Core Idea

Strong consistency guarantees that after a write completes, all subsequent reads will return that new value — no matter which server or replica you read from. It's as if there's only one copy of the data, always perfectly synchronized.

Think of it like a shared whiteboard in a classroom: when the teacher writes something, every student immediately sees the exact same content. There's no delay, no confusion, and no student sees an outdated version.

How It Works in Practice

In a strongly consistent system:

  • A write operation doesn't complete until all replicas have the updated value (or at least a quorum agrees)
  • Any read after that write sees the new data
  • The system coordinates carefully to ensure this single, linear order of operations

This is the strongest guarantee a distributed system can provide. It eliminates the confusing scenarios where different users see different values at the same time.

The Trade-off

Strong consistency requires coordination and waiting. If a replica is slow or unreachable, writes must wait or fail to maintain the guarantee. This relates directly to the CP (Consistency over Availability) systems we learned about in the CAP theorem — you sacrifice some availability to maintain perfect consistency.

Systems like distributed locking services and banking systems often choose strong consistency because correctness is non-negotiable.

Key Takeaway: Strong consistency guarantees all reads see the most recent write immediately, making distributed data behave like a single, synchronized copy — but it requires coordination that can impact availability and latency.