Course contentsShow
← Cloud and Infrastructure
Lesson 551 of 2,98910. API Gateway and Event-Driven PatternsFree lesson

Queue vs Topic Pattern

Distinguishing between point-to-point queues and publish-subscribe topics for different messaging scenarios.

Queue vs Topic Pattern

You'll learn: How to choose between point-to-point queues and publish-subscribe topics for different messaging scenarios.

The Core Difference

Both queues and topics handle asynchronous messaging, but they serve fundamentally different communication patterns:

Queues implement a point-to-point model. A message sent to a queue is consumed by exactly one receiver. Think of it like a task distribution system—once a worker picks up a task, no other worker processes it.

Topics implement a publish-subscribe model. A message published to a topic is delivered to all subscribed receivers. Think of it like a radio broadcast—every listener tuned to the station receives the same content.

When to Use Each Pattern

Queue Pattern

Use queues when you need:

  • Load distribution: Multiple workers processing tasks from a shared pool
  • Exactly-once processing: Each message handled by one consumer only
  • Work order preservation: First-in, first-out task execution
  • Example: Order processing, background job execution, batch data processing

Topic Pattern

Use topics when you need:

  • Fan-out notifications: Multiple systems reacting to the same event
  • Decoupled architectures: Publishers don't know who's listening
  • Parallel processing: Different services handling the same event differently
  • Example: User signup triggers email, analytics, and welcome workflow simultaneously

Combining Both Patterns

In practice, many architectures use both. A topic might distribute an event to multiple queues, each feeding a different processing service. This combines broadcast flexibility with load-balanced consumption.

Key Takeaway: Use queues for competitive consumption (one winner processes the message) and topics for collaborative consumption (all subscribers receive the message). The choice depends on whether your workload needs distribution or replication.