winston-azuretable/README.md

Azure Table Storage transport for Winston

[![NPM version](https://badge.fury.io/js/winston-azuretable.svg)](http://badge.fury.io/js/winston-azuretable)

winston-azuretable
==================

An Azure Table Storage transport for the [Winston](https://github.com/flatiron/winston) logging library.

Installation
------------

```bash
  $ npm install winston
  $ npm install winston-azuretable
```

Usage
-----
```
var winston = require('winston');
var azureLogger = require('winston-azuretable').AzureLogger

winston.add(azureLogger, options);
```

The transport accepts the following options:

* **useDevStorage**: Boolean denoting whether to use the Azure Storage Emulator (default: `false`)
* **account**: Azure Storage Account Name. If unset, will use `AZURE_STORAGE_ACCOUNT` environment variable.
* **key**: Azure Storage Account Key. If unset, will use  `AZURE_STORAGE_ACCESS_KEY` environment variable.
* **level**: logging level (default: `info`)
* **tableName**: name of table where logs will be outputted (default: `log`)
* **partitionKey**: table partition key (default: `process.env.NODE_ENV`)
* **silent**: Boolean flag indicating whether to suppress output to tables (default: `false`)
* **nestedMeta**: store metadata as a JSON document in `meta` column (default: `false`)

Table Entity
------------
Each log entry will create the following entity:

* **PartitionKey**: table partition key (default: `process.env.NODE_ENV`)
* **RowKey**: number of milliseconds since epoch
* **hostname**: hostname of operating system
* **pid**: node process id
* **level**: logging level
* **msg**: logging message
* **createdDateTime**: date log entry created
* when `nestedMeta` option is `false`:
  * *[metadata properties]*: creates associated property in entity for each given metadata property
* when `nestedMeta` option is `true`:
  * **meta**: JSON-encoded metadata properties


Inspirations/Alternatives
-------------------------
Inspired by [winston-skywriter](https://github.com/pofallon/winston-skywriter/). Differences in implementation include: 

* support for latest Azure Storage SDK
* dependency against Azure Storage SDK directly instead of an intermediary library (bluesky)
* row key is the number of milliseconds since epoch resulting in Azure table naturally having most recent log entry first
* implementation of Winston query method.