AWS Case Study- Yumch

The client wanted to have a pre-defined process for development and overall cost analysis if he migrated his whole hosting from another provider to AWS with continuous integration and continuous development process.

Scope

Web Development

Tech Stack

Amazon Elastic Beanstalk

Launch Site

Overview

Before migrating to AWS, our client came with a broken process and a codebase that was not properly maintained. In pursuit of a pre-defined process for development and overall cost analysis, he decided that if moving to AWS proved to be a wise choice, he would need a continuous integration and development process.

The Challenge

The client needed a defined process for Codebase development and maintenance, a systematic cost analysis for the old hosting provider and AWS, and a CI/CD pipeline for two different environments. They needed to migrate the entire existing server to AWS and create and configure three aspects, namely, RDS database as MySQL data source, Queue service from AWS, and mailing service from AWS.

The Solution

We first worked on defining a process for Codebase development and maintenance. At SJ Innovation, we follow a predefined process where we use agile methodology for every developmental step. We use a version control system from AWS known as Amazon CodeCommit to store/maintain the codebase and keep a track of changes taking place. Following this, a cost analysis of the services utilized for AWS was done while a comparison was made against the old hosting provider.

Next, we designed and developed CI/CD pipelines with different environments such as Dev, Stage, UAT, Production. Since it was a PHP Laravel site, we created a pipeline with services from AWS known as Amazon CodePipeline, Amazon Elastic Beanstalk, Amazon CodeCommit, and Amazon CodeDeploy among others. We then migrated the entire existing server to AWS so that every server-side configuration was in the right place with fully automated dependency management owing to custom scripts.

Depending on the needs of the application, we also moved the existing database which was in MySQL, and for AWS we used Amazon RDS Aurora which is a serverless relational database service from AWS. For a dynamic queue service system for the application, we suggested Amazon SQS [Simple Queue Service]. For a smooth mailing service for the application, we suggested Amazon SES.