Understanding Why Spring Cloud Uses HTTP for Remote Calls Instead of RPC

Spring Cloud services are built on an embedded Tomcat container that handles HTTP requests, allowing developers to exchange JSON data flexibly across platforms, which suits microservice architectures where a single server may serve H5, app, and mini‑program clients.

In contrast, RPC relies on TCP, requiring a three‑way handshake and binary serialization, making it less adaptable for frequently changing data structures and more suitable for stable, low‑level communication.

The article also explains the basics of socket connections, describing client and server sockets and the three‑step connection process (listen, request, confirm).

A simple Java HTTP server example demonstrates how browsers request URLs and receive HTML responses over HTTP.

Remote Procedure Call (RPC) is defined as a method for inter‑process communication that abstracts the details of remote invocation, allowing code to call functions on another machine as if they were local.

Key differences between RPC and HTTP/RESTful services are highlighted: RPC aims to make remote calls look like local calls and requires API‑level encapsulation, while HTTP offers more flexibility, language‑agnostic communication, and easier implementation.

The article lists advantages (RPC transparency, HTTP flexibility) and disadvantages (RPC’s need for API encapsulation), and provides guidance on choosing between them based on speed, complexity, and flexibility.

Future trends point toward microservice architectures that favor HTTP‑based RESTful services due to their independence and autonomy.