tracing-aka-logging help

[ximengyi]

I am learning to use axum
I cloned axum and went directly to /axum/examples/tracing-aka-logging to run cargo run
I can see:

2022-01-04T08:32:52.673597Z DEBUG example_tracing_aka_logging: listening on 127.0.0.1:3000

But I copied this example into my own project, but I can’t print this sentence. My code is exactly the same as the code in tracing-aka-logging.
But

tracing::debug!("listening on {}", addr);  //This code seems to be invalid.

This code cannot be printed to the console normally, I am confused ，Can you help me？

Below is the complete code and dependencies there is only one main.rs file

use axum::{
    body::Bytes,
    http::{HeaderMap, Request},
    response::{Html, Response},
    routing::get,
    Router,
};
use std::{net::SocketAddr, time::Duration};
use tower_http::{classify::ServerErrorsFailureClass, trace::TraceLayer};
use tracing::Span;

#[tokio::main]
async fn main() {
    // Set the RUST_LOG, if it hasn't been explicitly defined
    if std::env::var_os("RUST_LOG").is_none() {
        std::env::set_var(
            "RUST_LOG",
            "example_tracing_aka_logging=debug,tower_http=debug",
        )
    }
    tracing_subscriber::fmt::init();

    // build our application with a route
    let app = Router::new()
        .route("/", get(handler))
        // `TraceLayer` is provided by tower-http so you have to add that as a dependency.
        // It provides good defaults but is also very customizable.
        //
        // See https://docs.rs/tower-http/0.1.1/tower_http/trace/index.html for more details.
        .layer(TraceLayer::new_for_http())
        // If you want to customize the behavior using closures here is how
        //
        // This is just for demonstration, you don't need to add this middleware twice
        .layer(
            TraceLayer::new_for_http()
                .on_request(|_request: &Request<_>, _span: &Span| {
                    // ...
                })
                .on_response(|_response: &Response, _latency: Duration, _span: &Span| {
                    // ...
                })
                .on_body_chunk(|_chunk: &Bytes, _latency: Duration, _span: &Span| {
                    // ..
                })
                .on_eos(
                    |_trailers: Option<&HeaderMap>, _stream_duration: Duration, _span: &Span| {
                        // ...
                    },
                )
                .on_failure(
                    |_error: ServerErrorsFailureClass, _latency: Duration, _span: &Span| {
                        // ...
                    },
                ),
        );

    // run it
    let addr = SocketAddr::from(([127, 0, 0, 1], 3000));
    tracing::debug!("listening on {}", addr);
    tracing::info!("listening on {}", addr);
    axum::Server::bind(&addr)
        .serve(app.into_make_service())
        .await
        .unwrap();
}

async fn handler() -> Html<&'static str> {
     tracing::debug!("hello world");
    Html("<h1>Hello, World!</h1>")
}

[package]
name = "glimmer"
version = "0.1.0"
edition = "2021"
publish = false


[dependencies]
axum = { version = "0.4" }
tokio = { version = "1.0", features = ["full"] }
tracing = "0.1"
tracing-subscriber = { version="0.3", features = ["env-filter"] }
tower-http = { version = "0.2", features = ["trace"] }

[davidpdrsn]

Since your crate has a different name you need to update the log filter:

std::env::set_var(
    "RUST_LOG",
-   "example_tracing_aka_logging=debug,tower_http=debug",
+   "glimmer=debug,tower_http=debug",
)

[ximengyi]

It's okay thank you very much

[wemgl]

Hey @davidpdrsn, I came across this issue after running into the same problem. The solution you provided works, but I'd like to know where in the documentation I can find a description of an acceptable log filter configuration. I've already looked through https://github.com/tokio-rs/tower-http repo, and although the examples in that repo do use the crate's name in the log filter it didn't register to me that it had to be the crate's name, so I changed it to something else and saw that the log levels didn't change as I expected.

[davidpdrsn]

You'll wanna look at tracing's docs.

[wemgl]

Thanks, I was able to find it here: https://docs.rs/tracing-subscriber/0.2.12/tracing_subscriber/fmt/index.html#filtering-events-with-environment-variables

[1oglop1]

@davidpdrsn This is the most confusing thing I find as a beginner. The documentation is scattered across way too many places and it takes a lot of time to get something done or learn.
I'm comparing (Idk if it's a fair comparison) axum to Flask or Starlette where docs make more sense to me.

[davidpdrsn]

Do you have any suggestions for how to improve the situation? Keeping in mind that we can't simply document everything as part of axum.

[1oglop1]

I think that better interlinking between projects would help a lot. If docs suggested checking out related projects, eg. for this <concept> visit <X project>