Post

Using Protocol Buffers in Microservices

Using Protocol Buffers in Microservices

While JSON is human-readable and flexible, it’s often not the best choice for internal service-to-service communication.

Why Protobuf?

  • Smaller payload sizes
  • Faster (de)serialization
  • Strongly typed schema
  • Backward compatibility with versioning

I adopted Protobuf for communication between backend services on gRPC and REST. It required an initial tooling setup — codegen for Java and TypeScript clients, shared schemas in a Git repo — but the performance improvements were tangible.

Example schema:

1
2
3
4
5
6
7
8
syntax = "proto3";

message UserProfile {
  string id = 1;
  string name = 2;
  int32 age = 3;
}

JSON is still great for APIs consumed by humans (e.g., web/mobile clients), but for internal services, Protobuf is my new default.

This post is licensed under CC BY 4.0 by the author.