Remote Procedure Calls (RPC) Explained Simply: Concepts, Examples, and Use Cases

Modern software needs multiple machines to fulfill the full scope of its application. Software gets distributed across servers, cloud platforms, containers and multiple locations. Even despite this complexity, developers want systems to communicate remotely as easily as calling a local function. And this is the purpose of Remote Procedure Сalls (RPC).

In this article, we will explain what RPC is all about, how it works, and its examples and use cases.

What Is RPC (Remote Procedure Call)?

A Remote Procedure Call (RPC) is a communication technique that allows executing a function or procedure on another computer remotely as it would be done locally.

This is how it works:

• A program calls a function.
• The function is actually running on a different machine.
• The result is returned as if it were running on your computer.

RPC dissolves the complexity of networking for developers, which is palpable through formatting, data transferring, and handling of responses.

Why Developers Need RPC

RPC simplifies many tasks developers would otherwise have to do manually. For example:

• Sending network requests;
• Handling data;
• Handling timeouts and errors;
• Analyze responses;

Considering that RPC can cover these actions, it makes it much easier for developers to build, manage and maintain distributed systems.

How RPC Works (Step by Step)

Now, let’s get into the details of how RPC works. Here are the main steps:

1. Client calls a function, which resembles a local function call.
2. The function’s name and parameters are converted into a format suitable for network transmission.
3. The request is sent over the network.
4. The server receives the request, and executes the requested procedure.
5. The result is serialized and returned back to the client.
6. The client receives the response and converts it into a usable format.

An RPC Example

Imagine a function that adds two numbers: result = add(5, 3). With RPC, the add function runs on a remote server; the client sends 5 and 3; the server calculates 8; and the result is returned to a client.

Therefore, to a developer it feels like a normal function call.

Different types of RPC

There are three main types of RPC: synchronous, asynchronous, and one-way RPC.

In a synchronous RPC, the client waits for the server response, which can sometimes block the request execution. In asynchronous RPC, on the other hand, the client continues to work while waiting for a response, which can be better for high-performance systems.

One-way RPC enables a client to send a request and doesn’t wait for and expect a response.

Popular RPC Technologies

gRPC

gRPC technology:

• Was developed by Google;
• Uses HTTP/2 and Protocol Buffers;
• Is high-performance;
• Is commonly used in microservices;

JSON-RPC

JSON-RPC is:

• Lightweight;
• Using JSON for data exchange;
• Simple and easy-to-use no matter the language;

XML-RPC

XML-RPC is:

• An older RPC standard;
• Using XML;
• Less commonly used;

Thrift

Thrift:

• Was developed by Apache;
• Has cross-language support;
• Is used in large-scale distributed systems;

Common Use Cases for RPC

Microservices Communication

Services call each other efficiently with low latency.

Distributed Systems

Different components of a system run on separate machines but behave as one.

Cloud and Container Platforms

Internal communication between services in Kubernetes or cloud environments.

Database and Storage Systems

Remote data queries and operations.

Real-Time Systems

Gaming servers, financial systems, and messaging platforms.

Pros and Cons of RPC

Pros:

• Simple programming model;
• High performance;
• Language-neutral implementations;
• Clean abstraction for distributed systems;

Cons:

• Network failures must be handled;
• Tight coupling between client and server;
• Versioning can be complex;
• Debugging distributed calls is harder than local calls;