gRPC is an internal-service RPC framework that combines Protocol Buffers for binary serialization with HTTP/2 transport. It targets smaller messages, faster parsing, and efficient high-throughput communication between microservices. section 5.3.3
Its performance advantages come from several implementation choices:
- Binary payloads are roughly half the size of equivalent JSON messages. section 5.3.3
- Binary data parses faster than text, reducing latency and CPU usage across repeated service calls. section 5.3.3
- Explicit service contracts and cross-language compatibility address handwritten-client and multi-language service problems. section 5.3.3
Migration pressure
Netflix’s HTTP/1.1 internal RPC stack worked, but growth exposed unclear contracts, handwritten client code, and limited compatibility as teams adopted Node.js and Python alongside Java. section 5.3.3
Table 5.2 API approach comparison
|
|
REST
|
GraphQL
|
gRPC
|
|
Best for
|
Public APIs, cacheable resources
|
Diverse client needs, mobile apps
|
Internal services, high throughput
|
|
Payload size
|
Medium (JSON)
|
Medium (JSON)
|
Small (binary)
|
|
Caching
|
HTTP caching works naturally
|
Requires custom solutions
|
Requires custom solutions
|
|
Browser support
|
Native
|
Native
|
Requires proxy
|
|
Learning curve
|
Low
|
Medium
|
Medium-high
|
|
Debugging
|
Excellent
|
Good
|
Requires tooling
|
|
Round trips
|
One per resource
|
One for complex queries
|
One (streaming supported)
|
Compare gRPC with REST and GraphQL by payload size, caching, browser support, debugging, and round-trip behavior. section 5.3.3











