cache_feeder

This node package reads key-value pairs from a CSV file and inserts it into Redis using streams.

Application

Use this package if you want to set key-value pairs from a CSV file which contains huge number of rows (N_rows > 100,000). The script makes use of streams and a csv parser to read the file and make requests to cache in batches to reduce round trip time.

Why ?

At the start of a project or even in an ongoing project, sometimes it is required to insert a lot of key-value pairs into cache for the program to work properly. Inserting these (> 1 million) keys into cache by loading it into memory at once is not a good option and mostly results in app crashes. The better way to approach this problem is through streams.

Supported cache

1. Redis
2. Aerospike (In progress)

Installation

npm install --save cache_feeder

Github repository

https://github.com/anuj3918/cache_feeder.git

Code example

//  Setup options for library and cache
const options = {
    cacheConfig: {
        type: 'redis',
        host: '127.0.0.1',
        port: 6379
    },
    libSettings: {
        commandsPerBatch: 40,
        logTime: true,
        logProcesses: true,
        keyColName: 'key',
        valueColName: 'value'
    }
};

//  Requires cache_feeder and initialise it, cf here is a Promise object
const cf = require('./index')(options);

//  Starts setting key value pairs to cache, can use async-await here as well
cf('./MOCK_DATA.csv')
    .then(result => {
        console.log(result);
    });

Details of 'options' to be passed

Below are the default values of options with details

const options = {
    cacheConfig: {
        type: 'redis', // Type of cache
        host: '127.0.0.1',  //  Host of cache
        port: 6379  //  Port of cache
    },
    libSettings: {
        commandsPerBatch: 40, // Number of set commands to send in one batch request
        logTime: true,  //  Logs the total time taken on terminal
        logProcesses: true, //  Logs the tasks performed on terminal
        keyColName: 'key',  //  Name of column header to be set as key name
        valueColName: 'value' //  Name of column header to be set as value name
    }
};

Sample CSV file

first_name,email
anuj,[e-mail]
gupta,[e-mail]

To do

1. Support for Aerospike.
2. Adding more flexibility while creating connections to cache.
3. Log any failures while inserting.
4. Retry for only the failures keys encountered previously.
5. Load testing metrics to be added in Readme.md
6. Pausing and resuming stream while reconnecting to cache.

Contributions

You can add any of the above features into this package and create a pull request. For any further queries, write to [e-mail]