反向代理与CDN是否功能一致？Facebook是否使用反向代理？

Great questions—let’s break this down clearly, since these concepts are related but have key differences:

反向代理 vs CDN：核心异同 #

First, the overlap: both reverse proxies and CDNs act as an intermediary between clients and origin servers, and both can hide the origin server’s real IP address.

But the key distinctions are:

• A CDN is a distributed network of reverse proxies, focused primarily on content caching and low-latency delivery. For example, Facebook uses its CDN to cache images, videos, and static assets at edge locations worldwide—so when you load a photo, you pull it from a nearby edge node instead of Facebook’s core data centers. This reduces latency and offloads traffic from origin servers.
• A standard reverse proxy is usually a single or small cluster of servers, focused more on load balancing, request routing, security filtering, or acting as a gateway to backend services. It may not have global edge caching (though some can be configured with caching).

Does a reverse proxy perform redirects? #

It can, but redirects aren’t a core function of reverse proxies. Here’s how it works:

• If the origin server returns a 3xx redirect response, the reverse proxy will typically forward that response directly to the client.
• You can also configure a reverse proxy to initiate redirects on its own—for example, redirecting all HTTP requests to HTTPS, or sending users to a different URL based on their location or device.
• That said, many reverse proxy use cases (like load balancing) operate as transparent forwarders: the client sends a request, the proxy passes it to the origin, and returns the origin’s response without redirecting.

Is this mechanism for security purposes? #

Absolutely—hiding the origin server’s real IP is a critical security benefit, as it prevents attackers from directly targeting your backend infrastructure. But reverse proxies offer more security features beyond that:

• Integrate Web Application Firewalls (WAF) to block malicious requests (e.g., SQL injection, XSS attacks) before they reach the origin.
• Handle SSL/TLS encryption at the proxy layer, so origin servers don’t need to manage certificates or encryption overhead.
• Enforce rate limiting to prevent brute-force attacks or excessive traffic from overwhelming the origin.
• Block unwanted IP addresses or user agents to stop malicious crawlers or attackers.

Does Facebook use reverse proxies? #

Yes—extensively. Facebook’s infrastructure relies heavily on reverse proxies, including its CDN:

• Facebook’s Edge Network (its CDN) is a global distributed reverse proxy cluster. It caches static content, routes user requests to the nearest available node, and hides the origin servers’ real addresses.
• Beyond the CDN, Facebook uses reverse proxies in its core service layer to manage load balancing, route requests to the appropriate backend services, and enforce security policies. This helps keep their services stable, fast, and secure for billions of users.

内容的提问来源于stack exchange，提问作者Ben Druno