Forward proxies vs reverse proxies — what each one is actually for

A proxy, in general, is a middlebox that sits between two parties and relays requests on behalf of one of them. Both forward and reverse proxies do exactly that — the entire difference is which side they represent.

A forward proxy sits in front of one or more clients and forwards their requests outward. The server on the other end sees the proxy's identity, not the real client's — the proxy is acting for the client.

Common real examples:

A reverse proxy — already covered earlier in this domain, as the general role a load balancer plays — sits in front of one or more servers and forwards requests to whichever backend should actually handle them. The client sees the reverse proxy's identity, not the real backend server's — the proxy is acting for the server.

Common real examples:

The two terms get mixed up because the underlying mechanism — a middlebox relaying traffic — is identical. The only real question that tells them apart: whose identity is the proxy hiding from the other side? A forward proxy hides the client from the server. A reverse proxy hides the server (or servers) from the client. Once that one distinction is clear, "forward" and "reverse" stop being arbitrary labels and start describing exactly what each one does.

Recognizing which kind of proxy you're dealing with tells you where to look when something breaks. A request blocked by a corporate forward proxy is a client-network problem, unrelated to the destination server. A request that reaches the wrong backend, or gets a certificate error for a domain that doesn't match, usually points at reverse-proxy/load-balancer configuration instead — the same layer this domain's load balancer lesson already covered, just now correctly named as one specific kind of proxy rather than a separate concept.