A simple reverse proxy for use with Hyper and Tokio
Go to file
2024-02-28 16:30:08 +01:00
.github/workflows chore: update benches 2022-06-11 15:21:54 -03:00
benches chore: update benches 2022-06-11 15:21:54 -03:00
examples feat: static client (#27) 2022-05-01 15:36:09 -03:00
src Don't panic on upgrade errors 2024-02-28 16:30:08 +01:00
tests tests: refactor uris 2022-05-21 14:49:16 -03:00
.gitignore port lib to hyper 0.13 and future 0.3 2020-05-04 00:18:27 -03:00
.travis.yml Initial commit 2017-07-15 16:28:59 +10:00
Cargo.toml Upgrade to hyper 1.2.0 2024-02-27 22:34:41 +01:00
LICENSE Initial commit 2017-07-15 16:28:59 +10:00
README.md feat: static client (#27) 2022-05-01 15:36:09 -03:00

hyper-reverse-proxy

License CI docs version

A simple reverse proxy, to be used with Hyper.

The implementation ensures that Hop-by-hop headers are stripped correctly in both directions, and adds the client's IP address to a comma-space-separated list of forwarding addresses in the X-Forwarded-For header.

The implementation is based on Go's httputil.ReverseProxy.

Example

Add these dependencies to your Cargo.toml file.

[dependencies]
hyper-reverse-proxy = "?"
hyper = { version = "?", features = ["full"] }
tokio = { version = "?", features = ["full"] }
lazy_static = "?"
hyper-trust-dns = { version = "?", features = [
  "rustls-http2",
  "dnssec-ring",
  "dns-over-https-rustls",
  "rustls-webpki",
  "https-only"
] }

The following example will set up a reverse proxy listening on 127.0.0.1:13900, and will proxy these calls:

  • "/target/first" will be proxied to http://127.0.0.1:13901

  • "/target/second" will be proxied to http://127.0.0.1:13902

  • All other URLs will be handled by debug_request function, that will display request information.

use hyper::server::conn::AddrStream;
use hyper::service::{make_service_fn, service_fn};
use hyper::{Body, Request, Response, Server, StatusCode};
use hyper_reverse_proxy::ReverseProxy;
use hyper_trust_dns::{RustlsHttpsConnector, TrustDnsResolver};
use std::net::IpAddr;
use std::{convert::Infallible, net::SocketAddr};

lazy_static::lazy_static! {
    static ref  PROXY_CLIENT: ReverseProxy<RustlsHttpsConnector> = {
        ReverseProxy::new(
            hyper::Client::builder().build::<_, hyper::Body>(TrustDnsResolver::default().into_rustls_webpki_https_connector()),
        )
    };
}

fn debug_request(req: &Request<Body>) -> Result<Response<Body>, Infallible> {
    let body_str = format!("{:?}", req);
    Ok(Response::new(Body::from(body_str)))
}

async fn handle(client_ip: IpAddr, req: Request<Body>) -> Result<Response<Body>, Infallible> {
    if req.uri().path().starts_with("/target/first") {
        match PROXY_CLIENT.call(client_ip, "http://127.0.0.1:13901", req)
            .await
        {
            Ok(response) => {
                Ok(response)
            },
            Err(_error) => {
                Ok(Response::builder()
                .status(StatusCode::INTERNAL_SERVER_ERROR)
                .body(Body::empty())
                .unwrap())},
        }
    } else if req.uri().path().starts_with("/target/second") {
        match PROXY_CLIENT.call(client_ip, "http://127.0.0.1:13902", req)
            .await
        {
            Ok(response) => Ok(response),
            Err(_error) => Ok(Response::builder()
                .status(StatusCode::INTERNAL_SERVER_ERROR)
                .body(Body::empty())
                .unwrap()),
        }
    } else {
        debug_request(&req)
    }
}

#[tokio::main]
async fn main() {
    let bind_addr = "127.0.0.1:8000";
    let addr: SocketAddr = bind_addr.parse().expect("Could not parse ip:port.");

    let make_svc = make_service_fn(|conn: &AddrStream| {
        let remote_addr = conn.remote_addr().ip();
        async move { Ok::<_, Infallible>(service_fn(move |req| handle(remote_addr, req))) }
    });

    let server = Server::bind(&addr).serve(make_svc);

    println!("Running server on {:?}", addr);

    if let Err(e) = server.await {
        eprintln!("server error: {}", e);
    }
}

A word about Security

Handling outgoing requests can be a security nightmare. This crate does not control the client for the outgoing requests, as it needs to be supplied to the proxy call. The following chapters may give you an overview on how you can secure your client using the hyper-trust-dns crate.

You can see them being used in the example.

HTTPS

You should use a secure transport in order to know who you are talking to and so you can trust the connection. By default hyper-trust-dns enables the feature flag https-only which will panic if you supply a transport scheme which isn't https. It is a healthy default as it's not only you needing to trust the source but also everyone else seeing the content on unsecure connections.

ATTENTION: if you are running on a host with added certificates in your cert store, make sure to audit them in a interval, so neither old certificates nor malicious certificates are considered as valid by your client.

TLS 1.2

By default tls 1.2 is disabled in favor of tls 1.3, because many parts of tls 1.2 can be considered as attach friendly. As not yet all services support it tls 1.2 can be enabled via the rustls-tls-12 feature.

ATTENTION: make sure to audit the services you connect to on an interval

DNSSEC

As dns queries and entries aren't "trustworthy" by default from a security standpoint. DNSSEC adds a new cryptographic layer for verification. To enable it use the dnssec-ring feature.

HTTP/2

By default only rustlss http1 feature is enabled for dns queries. While http/3 might be just around the corner. http/2 support can be enabled using the rustls-http2 feature.

DoT & DoH

DoT and DoH provide you with a secure transport between you and your dns.

By default none of them are enabled. If you would like to enabled them, you can do so using the features doh and dot.

Recommendations:

  • If you need to monitor network activities in relation to accessed ports, use dot with the dns-over-rustls feature flag
  • If you are out in the wild and have no need to monitor based on ports, doh with the dns-over-https-rustls feature flag as it will blend in with other https traffic

It is highly recommended to use one of them.

Currently only includes dns queries as esni or ech is still in draft by the ietf