Envoy Proxy

Envoy Proxy Envoy

Cloud-native high-performance edge/middle/service proxy

★ 26,9K Reverse proxy Load balancer API gateway Network Observability

Homepage · Source code

Author: Envoy Project Authors · License: Apache-2.0

Version: 1.36.2 ·

Over Envoy

A high-performance C++ proxy built for cloud-native microservice deployments.

Envoy acts as a universal data plane for service mesh architectures, providing advanced routing, load balancing, and observability features. Created at Lyft and now a CNCF graduated project used by companies like Airbnb, Netflix, and Uber.

Envoy Architecture


Key Capabilities

  • High Performance: C++ implementation with minimal memory footprint
  • 🔀 Protocol Agnostic: Native HTTP/1.1, HTTP/2, HTTP/3, gRPC, and TCP support
  • 🎯 Intelligent Routing: Path-based routing, traffic splitting, and header manipulation
  • 🔄 Resilience Patterns: Automatic retries, circuit breakers, and timeout management
  • 📊 Rich Metrics: Built-in stats, distributed tracing, and access logging
  • 🔌 Dynamic Configuration: xDS APIs for runtime updates without restarts
  • 🛡️ Security First: TLS termination, mutual TLS, rate limiting, and authentication
  • 🌍 Multi-Protocol: WebSocket, MongoDB, Redis, Postgres wire protocols

Configuration Overview

This stack includes a basic static configuration. For your use case, you should:

  • Edit envoy.yaml to define your routing rules
  • Configure upstream clusters for your backend services
  • Set up listeners for your specific ports and protocols
  • Add filters for authentication, rate limiting, or other features

The default setup provides:

  • Admin dashboard at port 9901 for monitoring and debugging
  • Example HTTP listener at port 10000
  • Sample upstream service definition

Access Points

  • Admin Console: http://envoy.stack.localhost:9901 - View config, stats, and health
  • Proxy Endpoint: http://envoy.stack.localhost:10000 - Main traffic entry point

Resources