winston-s3-transport v1.2.0

Winston S3 Transport

Logs generated through Winston can be transferred to an S3 bucket using winston-s3-transport.

Installation

The easiest way to install winston-s3-transport is with npm.

npm install winston-s3-transport

Alternately, download the source.

git clone https://github.com/stegano/winston-s3-transport.git

Example

! The bucket path is created when the log is first created.

// Example - `src/utils/logger.ts`
import winston from "winston";
import S3Transport from "winston-s3-transport";
import { v4 as uuidv4 } from "uuid";
import { format } from "date-fns";

const s3Transport = new S3Transport({ s3ClientConfig: { region: "ap-northeast-2", }, s3TransportConfig: { bucket: "my-bucket", group: (logInfo: any) => { // Group logs with userId value and store them in memory. // If the 'userId' value does not exist, use the anonymous group. return logInfo?.message?.userId || "anonymous"; }, bucketPath: (group: string = "default") => { const date = new Date(); const timestamp = format(date, "yyyyMMddhhmmss"); const uuid = uuidv4(); // The bucket path in which the log is uploaded. // You can create a bucket path by combining group, timestamp, and uuid values. return /logs/${group}/${timestamp}/${uuid}.log; }, }, });

export const logger = winston.createLogger({ levels: winston.config.syslog.levels, format: winston.format.combine(winston.format.json()), transports: s3Transport, });

export default logger;

> Create log using winston in another module
```ts
// Example - another module
import logger from "src/utils/logger";
...
// Create a log containing the field `userId`
logger.info({ userId: 'user001', ....logs });

Configuration

s3ClientConfig

This library is internally using @aws-sdk/client-s3 to upload files to AWS S3.

• Please see AWSJavaScriptSDK/s3clientconfig

s3TransportConfig

bucket: string

• AWS S3 Bucket name

bucketPath: ((group: string) => string) | string

• AWS S3 Bucket path to upload log files

group?: (<T = any>(logInfo: T) => string) | string (default: "default")

• Group for logs classification.

dataUploadInterval?: number (default: 1000 * 20)

• Data upload interval(milliseconds)

fileRotationInterval?: number (default: 1000 * 60)

• File rotation interval(milliseconds)

maxDataSize?: number (default: 1000 1000 2)

• Max data size(byte)

Motivation

I made this so that it can be efficiently partitioned when storing log data in the S3 bucket. When you use vast amounts of S3 data in Athena, partitioned data can help you use the cost effectively.

Contributors ✨

Thanks goes to these wonderful people (emoji key):

This project follows the all-contributors specification. Contributions of any kind welcome!