Pub/Sub 入門

Welcome to Redis Pub/Sub! Pub/Sub, short for Publish/Subscribe, is a messaging pattern where senders (publishers) don't directly send messages to specific receivers (subscribers).

Instead, messages are sent to an intermediary, called a channel or topic. Subscribers then listen to these channels to receive messages.

In the Pub/Sub model, a publisher is the entity that creates and sends messages. Publishers don't need to know who (or if anyone) is listening.

• They focus only on sending messages to a specific channel.
• This makes the system flexible and scalable.

A subscriber is an entity interested in receiving messages. Subscribers express their interest by 'subscribing' to one or more specific channels.

• When a message is published to a channel, all active subscribers to that channel receive a copy.
• Subscribers also don't know who the publishers are.

A channel acts as a named conduit for messages. Publishers send messages to a channel, and subscribers receive messages from it.

Think of it like a specific TV channel: broadcasters (publishers) transmit content to it, and viewers (subscribers) tune in to watch.

Here's the basic message flow in Pub/Sub:

1. A subscriber tells the Pub/Sub system (like Redis) which channels it wants to listen to.
2. A publisher sends a message to a specific channel.
3. The Pub/Sub system immediately delivers that message to all subscribers currently listening to that channel.

One of the biggest advantages of Pub/Sub is decoupling. Publishers and subscribers don't directly communicate or even know about each other's existence.

This means you can easily add new publishers or subscribers without changing existing code, making your system more flexible and easier to maintain.

Pub/Sub is ideal for scenarios requiring real-time updates and notifications. Common examples include:

• Chat applications: Users send messages to a channel, and all chat participants receive them instantly.
• Live dashboards: Updates on stock prices or sensor data can be broadcast to many viewers.
• News feeds: New articles or events can be pushed to followers in real-time.

Unlike the traditional request/response model (where a client asks a server for data and waits for a reply), Pub/Sub is a push-based model.

• Request/Response: Client initiates communication, expects a direct reply.
• Pub/Sub: Publisher sends messages without expecting a reply; messages are pushed to subscribers.

Redis is an excellent choice for implementing Pub/Sub due to its speed and simplicity.

• It's an in-memory data store, ensuring very low latency for message delivery.
• Its Pub/Sub API is straightforward, making it easy to integrate into applications.
• Redis handles the channel management and message distribution efficiently.

What is the primary benefit of the Pub/Sub messaging model compared to direct communication?

In this lesson, you've learned the fundamental concepts of the Publish/Subscribe messaging model.

• We defined publishers, subscribers, and channels.
• You now understand how Pub/Sub enables decoupled communication and its suitability for real-time applications.
• We also touched upon why Redis is a great platform for Pub/Sub.

Next, we'll dive into the mechanics of how Redis implements Pub/Sub.