Rouille, Rust web server middleware VS quinn

Your CRUD web application server almost certainly doesn't need async Rust. Using a blocking HTTP server is not "might be a good idea", it simply is a good idea.

I recommend Rouille for this: https://github.com/tomaka/rouille. In case you are worried about performance, check the benchmark. Blocking Rouille is faster than builtin async server in Node.js.

As you say, the majority of the web ecosystem in Rust has moved to async - but if you’re happy to stray a bit from the beaten path then rouille might do the trick.

In strongly typed languages like Rust, composing smaller libraries is usually quite painless, so you don't need a large framework.

Personally for backend Rust I use rouille[0] for the server (it's very simple and async-free), askama[1] for compile-time HTML templates and (if a SPA is unavoidable, as that is of course always to be avoided if at all possible) yew[2] for client-side WASM.

Now this stack is what I like personally, but there are many options that you can combine, some more full-featured than others. Check out https://www.arewewebyet.org/ for a partial overview.

[0]: https://github.com/tomaka/rouille

[1]: https://github.com/djc/askama

[2]: https://yew.rs/

rouille

I'd like to put in a word for a simple, sync framework such as rouille. The compile times are much, much better, the number of dependencies is much smaller, the stuff it's built on (the standard library) is extensively tested and extremely reliable. Kernel context switches are slower than userspace thread scheduling, but not much slower, and as long as your services aren't just shoving bytes from one place to another (i.e. actually doing some computation) the time taken for a context switch vanishes into noise. A lot of benchmarks test how quickly a web service can move bytes, which (if your business logic is non-trivial) actually isn't the most critical factor.

I don't have any Rust-relevant experience here, but if I wanted to build a web server in Rust and was okay with "reasonable" performance, I'd probably give rouille a try first.

Rouille is fairly solid in my experience. Save the pain of async and spend it building software that works. Honestly with Rust's lack of GC you get predictable response times already.

How would you differentiate it from let's say Rouille ?

Since it lives on top of UDP, I believe all you need is SOCK_DGRAM, right? The rest of QUIC can be in a userspace library ergonomically designed for your programming language e.g. https://github.com/quinn-rs/quinn - and can interoperate with others who have made different choices.

Alternately, if you need even higher performance, DPDK gives the abstractions you'd need; see e.g. https://dl.acm.org/doi/abs/10.1145/3565477.3569154 on performance characteristics.

> Making things thread safe for runtime-agnostic utilities like WebSocket is yet another price we pay for making everything multi-threaded by default. The standard way of doing what I'm doing in my code above would be to spawn one of the loops on a separate background task, which could land on a separate thread, meaning we must do all that synchronization to manage reading and writing to a socket from different threads for no good reason.

Why so? Libraries like quinn[1] define "no IO" crate to define runtime-agnostic protocol implementation. In this way we won't suffer by forcing ourselves using synchronization primitives.

Also, IMO it's relatively easy to use Send-bounded future in non-Send(i.o.w. single-threaded) runtime environment, but it's almost impossible to do opposite. Ecosystem users can freely use single threaded async runtime, but ecosystem providers should not. If you want every users to only use single threaded runtime, it's a major loss for the Rust ecosystem.

Typechecked Send/Sync bounds are one of the holy grails that Rust provides. Albeit it's overkill to use multithreaded async runtimes for most users, we should not abandon them because it opens an opportunity for high-end users who might seek Rust for their high-performance backends.

[1]: https://github.com/quinn-rs/quinn

quicssh-rs is quicssh rust implementation. It is based on quinn and tokio

Regarding Quinn, I had a blast this week resurrecting an old PR. Looking forward to the next!

By studying u/djcu/hachyderm.io (and others!) excellent work in Quinn, doing a sans I/O implementation of QUIC https://github.com/quinn-rs/quinn we have a way forward.

QUIC isn't web/wasm-compatible because of https://github.com/quinn-rs/quinn/issues/1388, so durian wouldn't either since it's built on top of it.

Quinn, as in the implementation of QUIC? https://github.com/quinn-rs/quinn

You are right, I am planning to switch the transport to UDP + quic using the awesome QUINN library, https://github.com/quinn-rs/quinn .

Hi, I am new to the community, I just started learning rust and created a secure UDP tunnel based on the Quinn library, thanks to Quinn, I didn't need to go into the detail of the QUIC protocol and quickly created a UDP tunnel, and thanks to the BBR congestion control algorithm it uses, the tunnel performs quite well with lousy and long fat network, I didn't do any benchmark, but it performs a lot better (higher throughput with LFN) than most of other TCP tunnel implementations I used before.