MVF Metrics

Usage

Provides a consistent way to log metrics in javascript applications.

To install the package

Run npm install @mvf/metrics

Configuration

Set the following environment variables in your project

• DATADOG_HOST or DATADOG_HOST_ENVVAR if the datadog host is stored in another environment variable, for example, if datadog host is assigned to HOST env variable then you can set DATADOG_HOST_ENVVAR=HOST which will then load datadog host from HOST environment variable.
• DATADOG_PORT or defaults to 8125 if not set.
• DATADOG_MOCK should be set to true in development and testing environments. This will setup hot-shot Client in a mock mode.
• DATADOG_PROJECT_NAME
• DATADOG_SERVICE_NAME

More on DATADOG_PROJECT_NAME

This is the system name of your project, e.g. abbey, mercury, leads, etc. DATADOG_PROJECT_NAME is used as the default prefix of your metric name, for example if DATADOG_PROJECT_NAME is set to

DATADOG_PROJECT_NAME=abbey

in your environment and you run

histogram('my.metric', 5);

or any other metric function then the full metric name sent to datadog will be abbey.my.metric.

More on DATADOG_SERVICE_NAME

This is the type of service in your project, e.g. web, api, cron, consumer, worker, etc. DATADOG_SERVICE_NAME is attached to every metric as a tag and should be used to group metrics by the type of service, for example if DATADOG_SERVICE_NAME is set to

DATADOG_SERVICE_NAME=cron

in your environment and you run

histogram('my.metric', 5);

or any other metric function then that metric will have service:cron tag.

Decrement

To decrement index metric by 5 use the following

import { decrement } from '@mvf/metrics';

decrement('index', 5);

Gauge

To gauge index metric to 50 use the following

import { gauge } from '@mvf/metrics';

gauge('index', 5);

Histogram

To gauge index metric to 50 use the following

import { histogram } from '@mvf/metrics';

histogram('index', 5);

Increment

To increment index metric by 5 use the following

import { increment } from '@mvf/metrics';

increment('index', 5);

Time

This is used to time how long something took to run. There are 3 ways this function can be used.

Call with explicit value

To log 1000 milliseconds for calculateFavorites metric use the following

import { time } from '@mvf/metrics';

const milliseconds = 1000;
time('calculateFavorites', milliseconds);

Call with callback

To count how long something takes to run and log that time use the callback approach

import { time } from '@mvf/metrics';

const calculateFavorites = () => {
  // do something
};

time('calculateFavorites', calculateFavorites);

Call with async callback

If your callback is asynchronous you can use async function and await

import { time } from '@mvf/metrics';

const calculateFavorites = async () => {
  // do something
};

await time('calculateFavorites', calculateFavorites);

Return value from callback

Any value returned from the callback will be return by the time function, e.g.

sync callback

const data = time('calculateFavorites', () => 'data');

async callback

const data = await time('calculateFavorites', async () => 'data');

Unique

To log 50 for currentIndex metric use the following

import { unique } from '@mvf/metrics';

unique('currentIndex', 50);

Contributing

Setup

• Run make to build the container
• Run make shell to enter the container
• Run npm install to install dependencies

Refer to package.json for commands

After merging

After you have merged a PR to master, you need to rebuild and publish your changes.

1. Checkout master git checkout master && git pull
2. Use one of the following make publish commands to publish changes:
   • make publish kind=patch - Use this if your change is a bug fix and is backwards compatible.
   • make publish kind=minor - Use this if your change adds new functionality and is backwards compatible.
   • make publish kind=major - Use this if your change is not backwards compatible.