Skip to content
This repository has been archived by the owner on Oct 1, 2024. It is now read-only.

Latest commit

 

History

History
109 lines (68 loc) · 2.79 KB

README.md

File metadata and controls

109 lines (68 loc) · 2.79 KB

@shopify/koa-metrics

Caution

@shopify/koa-metrics is deprecated.

Shopifolk, see Shopify/quilt-internal for information on the latest packages available for use internally.

Build Status Build Status License: MIT npm version

Opinionated performance metric tracking for Koa, implemented with DogStatsD.

Installation

yarn add @shopify/koa-metrics

Usage

import Koa from 'koa';
import metrics from '@shopify/koa-metrics';

const app = new Koa();

app.use(
  metrics({
    prefix: 'AppName',
    host: 'some-statsd-host.com:8125',
  }),
);

Options

The passed in options object adheres to the following API:

prefix (required)

The global StatsD metric name prefix; should be provided in PascalCase.

host (required)

The url for the StatsD host; should be provided in the format: hostname:port.

API

An instance of the Metrics object will be available on ctx.metrics further down in the middleware stack.

Metrics

.distribution(name: string, value: number, sampleRate?: number, tags?: Tags)

Sends a distribution command with the specified value in milliseconds.

.initTimer(): Timer

Returns a new Timer started at the current process.hrtime()

Timer

.stop(): number

Returns the time, in milliseconds, since the Timer was created.

Tags

tags are an object keyed by the name of the corresponding tag. For example:

{
  name: value,
  name2: value2
}

Intelligent Defaults

The global metric name prefix is provided through the options object.

This package automatically provides performance metrics for HTTP requests.

Standard tags

  • path
  • request_method
  • response_code
  • response_type (eg. 2xx, 3xx, ...)

Default Metrics

request_time

Time to complete a request, from the application perspective.

request_queuing_time

Time before a request actually started being processed.

This metric is emitted when the application start processing a request. It relies on the presence of a header X-Request-Start set by the first HTTP hop.

request_content_length

This metric is based on the response header Content-Length. Some responses don't provide this header (chunked encoding); in those cases, this will not be reported.