openst / storage Goto Github PK

PROBLEM : Making sure a coding style is properly followed is a big undertaking, it often slips through the cracks leaving the work inconsistent across repos.

SOLUTION
We came across Prettier https://prettier.io/ that automatically style the code. This will help the codebase be uniform and easier to read alongwith saving good amount of our time.

Prettier supports:

JavaScript, including ES2017
JSX
Flow
TypeScript
CSS, Less, and SCSS
JSON
GraphQL
Markdown, including GFM
YAML
It removes all original styling and ensures the outputted code conforms to a consistent style. (See this post)

Problem : Environment variables are a hard dependency. These lead to problems when multiple instances of OpenST-Storage are needed, each having connection to different cache instances (for example). Thus making scaling not possible.

Solution : Remove dependency on environment variables. How do we do that - Instead of reading the environment variables, we depend on config object passed in the OpenST Storage constructor. Thus 2 objects of OpenST Storage can have totally different config strategies and thus will not interfere amongst each other.

This will enable us to instantiate multiple instances of "openst-storage" based on an applications infrastructure requirements.

The project uses outdated dependencies (older versions) which have publically reported vulnerabilities. The versions should be updated and tested so that the existing code base doesn't break.

We should expose the Base model of DynamoDB so that other projects can use Sharded Base Model for their purposes. This would improve the performance significantly.

We should make use of Amazon DynamoDB Accelerator (DAX) to speed up certain actions, such as 'Batch Get', 'Batch Write', 'Update Item', etc. DAX delivers up to a 10x performance improvement – from milliseconds to microseconds – even at millions of requests per second. More information on DAX can be found here.

This would help increase the performance of certain operations.