Skip to content

imqueue/async-logger

Repository files navigation

@imqueue/async-logger

Configurable async logger over winston for @imqueue services.

Install

npm i --save @imqueue/async-logger

Usage

There are two actual ways of using async logger.

  1. Rely on singleton instance produced by this library, which is configured by environment variables (see Configuration section below)

    In this case as simple as

    import logger from '@imqueue/async-async-logger';
    
    serviceOptions.logger = logger;
  2. Instantiate and configure async logger programmatically:

    import { Logger } from '@imqueue/async-async-logger';
    
    const { name, version } = require('./package.json');
    
    const logger = new Logger({
        transports: [{
            type: 'http',
            options: {
                ssl: true,
                port: 443,
                path: '/v1/input/<YOUR_API_KEY>',
                host: 'http-intake.logs.datadoghq.com',
                headers: {
                    'Content-Type': 'application/json',
                },
            },
            enabled: true,
        }],
        metadata: {
            ddsource: `${ name } ${ version }`,
            ddtags: 'env: dev',
            hostname: 'localhost'
        },
    });
    
    serviceOptions.logger = logger;

Configuration

Logger can be configured via environment variables. It can be easily integrated with any remote services. Here we will example configuration based on assumption to connect with Datadog remote service.

So, basically there are two basic configuration options available:

export LOGGER_TRANSPORTS='[]'
export LOGGER_METADATA='{}'

Both of them are simply JSON strings referring to replicate corresponding winston logger settings. Both of them can accept string tags %name, %version for dynamic setting of these values from a current service package, if needed.

  1. To configure HTTP transport follow to pass such object to LOGGER_TRANSPORTS array:
{
    "type": string,
    "options": {
        "ssl": boolean,
        "port": number,
        "path": string,
        "host": string,
        "headers": object
    },
    "enabled": boolean
}
  1. To configure File transport
{
    "type": string,
    "options": {
        "filename": string,
        "dirname": string,
        "options": object,
        "maxsize": number,
        "zippedArchive": boolean,
        "maxFiles": number,
        "eol": string,
        "tailable": boolean
    },
    "enabled": boolean
}

EXAMPLE FOR DATADOG:

export LOGGER_TRANSPORTS='[{"type":"http","options":{"ssl":true,"port":443,"path":"/v1/input/[DATADOG_API_KEY]","host":"http-intake.logs.datadoghq.com","headers":{"Content-Type":"application/json"}}, "enabled": true }]'

where DATADOG_API_KEY should be replaced with an actual API key.

  1. To configure metadata consider the following LOGGER_METADATA object:
{
    "ddsource": string,
    "ddtags": string,
    "hostname": string
}

EXAMPLE FOR DATADOG:

export LOGGER_METADATA='{"ddsource":"%name %version","ddtags":"env: dev","hostname":"localhost"}'

Lets See It Human-Readable:

# example of transports config
export LOGGER_TRANSPORTS='[{
    "type": "http",
    "options": {
        "ssl": true,
        "port": 443,
        "path": "/v1/input/<YOUR_API_KEY>",
        "host": "http-intake.logs.datadoghq.com",
        "headers": {
            "Content-Type": "application/json"
        }
    },
    "enabled": true
}, {
    "type": "file",
    "options": {
        "filename": "logs.log",
        "dirname": "/home/usr/logs",
        "options": object,
        "zippedArchive": false,
    }
    "enabled": true
}]'
# example of metadata for datadog
export LOGGER_METADATA='{
    "ddsource": "%name %version",
    "ddtags": "env: dev",
    "hostname": "localhost"
}';

Contributing

Any contributions are greatly appreciated. Feel free to fork, propose PRs, open issues, do whatever you think may be helpful to this project. PRs which passes all tests and do not brake tslint rules are first-class candidates to be accepted!

License

ISC

Happy Coding!