nntp-proxy

High-throughput NNTP proxy written in Rust.

nntp-proxy sits between NNTP clients and one or more backend servers. It gives you a single local endpoint while handling backend selection, pooling, authentication, optional caching, and metrics in one place.

What it does

• Shares multiple backend servers across multiple clients
• Pools and reuses backend connections instead of having every client open its own
• Supports backend auth, outbound TLS, health checks, and connection limits
• Tracks article availability and can optionally cache article bodies
• Runs as one binary: nntp-proxy

Client-facing connections are plain NNTP only. TLS support is for outbound backend connections, not for the local listener.

Performance snapshot

Historical published Ryzen 9 results saw the 1/1/1 shape (1 client, 1 proxy thread, 1 backend connection) reach 4797.65 MiB/s in the 10GiB matrix, while separate 100GiB spot checks measured 4712.73 MiB/s mean over 10 runs. See CHANGELOG.md for the recorded numbers and docs/development.md for the current benchmark workflow.

Current status

Hybrid is the default routing mode. per-command and stateful remain available as explicit modes. Details are in docs/operator/runtime-and-routing.md.

Quick start

cargo build --release
cp config.minimal.toml config.toml
./target/release/nntp-proxy --config config.toml

Then edit config.toml so [[servers]] points at a real backend and connect your NNTP client to localhost:8119 unless you changed [proxy].port.

Read next

• Getting started
• Configuration
• Caching
• Operations
• Development

Example configs

• config.minimal.toml
• config.example.toml
• config.full.toml
• config.cache.toml

License

MIT