Apache Thrift

RPC framework and interface definition language for building cross-language services. You define data types and service interfaces in a Thrift IDL file, then the code generator produces client and server stubs for dozens of languages. Data serialization and transport are abstracted into a pluggable stack, so services written in different runtimes can communicate without manual serialization logic or boilerplate.

Originally developed at Facebook and donated to the Apache Software Foundation. Used in production by Meta, Evernote, Pinterest, Uber, Cloudera, and Siemens, and powers Apache projects such as HBase, Hadoop, Parquet, and Storm. Apache Thrift differentiates from gRPC and Protocol Buffers by supporting 28 programming languages and by design for non-atomic version changes, letting you roll out server updates while older clients remain compatible.

Core capabilities:

• IDL-driven code generation across 28 languages including C++, Java, Python, Go, Rust, Swift, and Node.js
• Pluggable transports: TCP sockets, HTTP, TLS, Unix domain sockets, memory, pipes
• Multiple serialization protocols: Binary, Compact, and JSON
• Backward-compatible evolution so clients and servers can upgrade independently

Microservices needing inter-language RPC, polyglot backends where Python frontends call Java services, and data pipelines integrating Hadoop or HBase benefit from a single IDL as the source of truth. The compiler output integrates with build systems, and libraries ship via common package managers.