actix-web-metrics

Metrics.rs integration for actix-web.

By default two metrics are tracked:

• http_requests_total (labels: endpoint, method, status): the total number of HTTP requests handled by the actix HttpServer.

• http_requests_duration_seconds (labels: endpoint, method, status): the request duration for all HTTP requests handled by the actix HttpServer.

Usage

First add actix-web-metrics to your Cargo.toml:

[dependencies]
actix-web-metrics = "x.x.x"

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

use std::collections::HashMap;

use actix_web::{web, App, HttpResponse, HttpServer};
use actix_web_metrics::{ActixWebMetrics, ActixWebMetricsBuilder};
use metrics_exporter_prometheus::PrometheusBuilder;

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

#[actix_web::main]
async fn main() -> std::io::Result<()> {
    // Register a metrics exporter.
    // In this case we will just expose a Prometheus metrics endpoint on localhost:9000/metrics
    //
    // You can change this to another exporter based on your needs.
    // See https://github.com/metrics-rs/metrics for more info.
    PrometheusBuilder::new().install().unwrap();
    // Configure & build the Actix-Web middleware layer
    let metrics = ActixWebMetricsBuilder::new()
        .build()
        .unwrap();

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

In the example above we are using the PrometheusBuilder from the metrics-exporter-prometheus crate which exposes the metrics via an HTTP endpoint.

A call to the localhost:9000/metrics endpoint will expose your metrics:

$ curl http://localhost:9000/metrics
# HELP http_requests_total Total number of HTTP requests
# TYPE http_requests_total counter
http_requests_total{endpoint="/health",method="GET",status="200",label1="value1"} 1

# HELP http_requests_duration_seconds HTTP request duration in seconds for all requests
# TYPE http_requests_duration_seconds summary
http_requests_duration_seconds{endpoint="/health",method="GET",status="200",label1="value1",quantile="0"} 0.000302807
http_requests_duration_seconds{endpoint="/health",method="GET",status="200",label1="value1",quantile="0.5"} 0.00030278122831198045
http_requests_duration_seconds{endpoint="/health",method="GET",status="200",label1="value1",quantile="0.9"} 0.00030278122831198045
http_requests_duration_seconds{endpoint="/health",method="GET",status="200",label1="value1",quantile="0.95"} 0.00030278122831198045
http_requests_duration_seconds{endpoint="/health",method="GET",status="200",label1="value1",quantile="0.99"} 0.00030278122831198045
http_requests_duration_seconds{endpoint="/health",method="GET",status="200",label1="value1",quantile="0.999"} 0.00030278122831198045
http_requests_duration_seconds{endpoint="/health",method="GET",status="200",label1="value1",quantile="1"} 0.000302807
http_requests_duration_seconds_sum{endpoint="/health",method="GET",status="200",label1="value1"} 0.000302807
http_requests_duration_seconds_count{endpoint="/health",method="GET",status="200",label1="value1"} 1

NOTE: There are 2 important things to note:

• The metrics-exporter-prometheus crate can be swapped for another metrics.rs compatible exporter.
• The endpoint exposed by metrics-exporter-prometheus is not part of the actix web http server.

Features

Custom metrics

The metrics.rs crate provides macros for custom metrics. This crate does not interfere with that functionality.

use actix_web::{web, App, HttpResponse, HttpServer};
use actix_web_metrics::{ActixWebMetrics, ActixWebMetricsBuilder};
use metrics::counter;

async fn health() -> HttpResponse {
    counter!("my_custom_counter").increment(1);
    HttpResponse::Ok().finish()
}

#[actix_web::main]
async fn main() -> std::io::Result<()> {
    let metrics = ActixWebMetricsBuilder::new()
        .build()
        .unwrap();

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

Configurable routes pattern cardinality

Let's say you have on your app a route to fetch posts by language and by slug GET /posts/{language}/{slug}. By default, actix-web-metrics will provide metrics for the whole route with the label endpoint set to the pattern /posts/{language}/{slug}. This is great but you cannot differentiate metrics across languages (as there is only a limited set of them). Actix-web-metrics can be configured to allow for more cardinality on some route params.

For that you need to add a middleware to pass some extensions data, specifically the MetricsConfig struct that contains the list of params you want to keep cardinality on.

use actix_web::{dev::Service, web, HttpMessage, HttpResponse};
use actix_web_metrics::ActixWebMetricsExtension;

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

web::resource("/posts/{language}/{slug}")
    .wrap_fn(|req, srv| {
        req.extensions_mut().insert::<ActixWebMetricsExtension>(
            ActixWebMetricsExtension { cardinality_keep_params: vec!["language".to_string()] }
        );
        srv.call(req)
    })
    .route(web::get().to(handler));

See the full example with_cardinality_on_params.rs.

Configurable metric names

If you want to rename the default metrics, you can use ActixWebMetricsConfig to do so.

use actix_web_metrics::{ActixWebMetricsBuilder, ActixWebMetricsConfig};

ActixWebMetricsBuilder::new()
    .metrics_config(
        ActixWebMetricsConfig::default()
        .http_requests_duration_seconds_name("my_http_request_duration")
        .http_requests_duration_seconds_name("my_http_requests_duration_seconds"),
    )
    .build()
    .unwrap();

See full example configuring_default_metrics.rs.

Masking unmatched requests

By default, if a request path is not matched to an Actix Web route, it will be masked as UNKNOWN. This is useful to avoid producing lots of useless metrics due to bots or malicious actors.

This can be configured in the following ways:

• mask_unmatched_patterns() can be used to change the endpoint label to something other than UNKNOWN.
• disable_unmatched_pattern_masking() can be used to disable this masking functionality.

use actix_web_metrics::ActixWebMetricsBuilder;

ActixWebMetricsBuilder::new()
    .mask_unmatched_patterns("UNMATCHED")
    // or .disable_unmatched_pattern_masking()
    .build()
    .unwrap();

The above will convert all /<nonexistent-path> into UNMATCHED:

http_requests_duration_seconds_sum{endpoint="/favicon.ico",method="GET",status="400"} 0.000424898

becomes

http_requests_duration_seconds_sum{endpoint="UNMATCHED",method="GET",status="400"} 0.000424898

Motivations

actix-web-metrics is heavily inspired (and forked from) actix-web-prom. Special thanks to @nlopes for their excellent work on actix-web-prom.

This crate replaces the underlying metrics implementation from the prometheus crate with metrics.rs.

The reasons for doing this are as followed:

• The metrics.rs ecosystem provides more ergonomic ways to instrument applications than the raw prometheus client.
• metrics.rs provides more customizable ways to export metrics.
• The future of the prometheus crate is uncertain (see https://github.com/tikv/rust-prometheus/issues/530)