actix-web-metrics-mw

Generic middleware library for actix-web metrics aggregation, can send to various outlets.

![Build Status](https://github.com/Igosuki/actix-web-metrics-mw/workflows/build.yml/badge.svg

Metrics middleware instrumentation for actix-web using the metrics-rs crate .

By default two metrics are tracked (this assumes the namespace actix_web_metrics_mw):

Available exporters :

• Statsd : uses a statsd rust client that buffers metrics through UDP, dogstats format is supported for metric labels

Default metrics :

• http_requests_total (labels: endpoint, method, status): request counter for each endpoint and method.
• http_requests_duration (labels: endpoint, method, status): histogram of request durations for each endpoint.

Dependencies

• actix 2
• futures 0.3
• metrics 0.12

Issues

Please feel free to submit issues for evolutions you feel are necessary.

Usage

First add actix_web_metrics_mw to your Cargo.toml:

[dependencies]
actix_web_metrics_mw = "0.2.0"

You then instantiate the prometheus middleware and pass it to .wrap():

use actix_web::{web, App, HttpResponse, HttpServer};
use actix_web_metrics_mw::Metrics;

fn health() -> HttpResponse {
    HttpResponse::Ok().finish()
}

fn main() -> std::io::Result<()> {
    let metrics = Metrics::new("/metrics", "actix_web_mw_test");
    HttpServer::new(move || {
        App::new()
            .wrap(metrics.clone())
            .service(web::resource("/health").to(health))
    })
    .bind("127.0.0.1:8080")?
    .run();
    Ok(())
}

Using the above as an example, a few things are worth mentioning:

• api is the metrics namespace
• /metrics will be auto exposed (GET requests only)

A call to the /metrics endpoint will expose your metrics:

$ curl http://localhost:8080/metrics
{"http_requests_total":"1570","http_requests_duration":"[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]"}

Custom metrics

You can instantiate Metrics and then use its sink to register your custom metric.

You can also use the metrics library macros or the entire metrics runtime to add new metrics and labels as suit your needs.

use actix_web::{web, App, HttpResponse, HttpServer};
use actix_web_metrics_mw::Metrics;

fn health() -> HttpResponse {
    counter!("endpoint.method.status", 1);
    HttpResponse::Ok().finish()
}

fn main() -> std::io::Result<()> {
    let metrics = Metrics::new("/metrics", "actix_web_mw_test");

    HttpServer::new(move || {
        App::new()
            .wrap(metrics.clone())
            .service(web::resource("/health").to(health))
    })
    .bind("127.0.0.1:8080")?
    .run();
    Ok(())
}

Live functional testing

Use the docker-compose file. Actual result :

Special Thanks

• The middleware integration is influenced by the work in nlopes/actix-web-prom.