设计实时聊天系统
逐步了解实时聊天应用的系统设计，涵盖连接、消息传递、存储和扩展。
设计实时聊天系统 是 CoddyKit 上的免费 System Design Basics for Backend Developers 课时。 这是第 4 节课，共 4 节。 你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。 这是 System Design Basics for Backend Developers 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 System Design Basics for Backend Developers 课程共包含 4 节课。
本课时的部分内容尚未翻译，以英文显示。
The Problem
Design a chat system like WhatsApp or Slack: users send messages that appear on recipients' devices instantly, with history, online presence, and delivery receipts — at the scale of millions of concurrent users.
Requirements
Clarify scope first:
- Functional: 1:1 and group chat, message history, presence, delivery/read receipts
- Non-functional: low latency, high availability, ordered delivery, horizontal scale
Why Not Plain HTTP?
Classic HTTP is request-response — the server cannot push. Polling wastes resources and adds latency. Real-time chat needs a persistent, bidirectional connection.
WebSockets
WebSockets upgrade an HTTP connection into a long-lived, full-duplex channel. The server can push a message the instant it arrives. This is the backbone of real-time chat.
GET /chat HTTP/1.1
Upgrade: websocket
Connection: Upgrade
# After upgrade: server can push frames anytimeConnection Servers
A fleet of connection servers holds the open WebSockets. Since a sender and recipient may be connected to different servers, you need a way to route a message from one server to another.
Routing Between Servers
A presence/session store (e.g. Redis) maps each user to the connection server holding their socket. A message broker or pub/sub forwards messages to the right server so it can push to the recipient.
user_to_server = {
'alice': 'conn-3',
'bob': 'conn-7'
}
# alice -> bob: route message to conn-7Storing Messages
Chat is write-heavy with huge volume. A wide-column store like Cassandra suits it: partition by conversation id, cluster by timestamp, so fetching recent history in order is fast.
PRIMARY KEY ((conversation_id), message_ts)
WITH CLUSTERING ORDER BY (message_ts DESC)Delivery Guarantees
Use a per-user inbox queue and acknowledgements. Mark messages sent, delivered, and read. If a recipient is offline, queue the message and deliver when they reconnect; retries plus message IDs keep it idempotent.
Ordering
Messages must appear in a consistent order. Attach a monotonic sequence or timestamp per conversation. Clients sort by it, so even out-of-order network delivery is corrected on display.
import time
base = int(time.time() * 1000)
seq = [base, base + 1, base + 2]
print('ordered message ids:', seq)Group Chat at Scale
Group messages fan out to every member. For small groups, push directly; for large ones, write once to the conversation and let members pull, or fan out asynchronously via the broker to avoid a write storm.
Putting It Together
The full picture: clients hold WebSockets to connection servers, a presence store routes via pub/sub, messages persist in a wide-column store, and queues plus receipts handle offline delivery and ordering. Each piece scales horizontally.
Quick Check
Test your understanding of real-time chat design.
Recap
You designed a real-time chat system:
- WebSockets give persistent bidirectional connections
- A presence store plus pub/sub routes messages between connection servers
- A wide-column store holds ordered history
- Queues and receipts handle offline delivery; ordering uses per-conversation sequences
用 AI 导师学习 System Design Basics for Backend Developers — 免费
在浏览器中编写并运行真实代码，获得全天候 AI 导师的即时帮助，并在网页或应用中继续学习。
- 课程
- 12
- 课程
- 48
常见问题解答
「设计实时聊天系统」课时是免费的吗？
是的 — 「设计实时聊天系统」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 System Design Basics for Backend Developers 课程的其余内容，请升级到 CoddyKit PRO。 System Design Basics for Backend Developers 课程共包含 4 节课。
「设计实时聊天系统」这节课中我会学到什么？
逐步了解实时聊天应用的系统设计，涵盖连接、消息传递、存储和扩展。 你通过在浏览器中直接运行的动手代码来练习 System Design Basics for Backend Developers，全天候 AI 导师会在你学习这节课的过程中回答你的问题。
学习 System Design Basics for Backend Developers 需要有经验吗？
无需任何先前经验。CoddyKit 上的 System Design Basics for Backend Developers 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。 这是第 4 节课，共 4 节。
「设计实时聊天系统」课时需要多长时间？
大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。
我能在这节 System Design Basics for Backend Developers 课中编写并运行代码吗？
能。每节 System Design Basics for Backend Developers 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。
此课程中的所有课时
- 设计 URL 缩短服务
- 构建社交媒体信息流
- 扩展电子商务平台
- 设计实时聊天系统