CLIENT OVERVIEW

The client is a well known player in EdTech Space providing learning and testing services to students aspiring to get admission in to undergraduate courses at leading universities across the globe.

CHALLENGE

The client runs several responsive web applications and mobile apps. With massive response for some of their new offerings, their customer base expanded rapidly from 2,000 to 250,000 in a short span of time. This rapid scaling up exposed many inherent issues in the design of the application and the supporting infrastructure and resulted in a range of scalability issues. We were hired to take a comprehensive look at the client’s infrastructure on AWS and architect a scalable infrastructure to support scaling up to 2 million end users and also suggest improvements to the application design.

ENGAGEMENT SUMMARY & SOLUTION

We worked closely with the client’s application team and identified the following improvements needed at the application level:

• Enhancing the application to make it suitable for horizontal scaling
• Rationalizing the use of third party APIs to reduce the total impact on performance
• Use of caching to reduce load on the web/api servers and to improve performance
• Moving to asynchronous integrations and loose coupling between components

Cloud migration for scalability

We helped a media house refactor the product to make it more cloud optimized.
Here are some of the optimizations:

• All scheduled jobs migrated to serverless services
• All static assets moved to object storage and placed behind CDN
• DBs migrated managed DBs for better manageability
• Compute services optimized with horizontal scalability
• Health checks with notification to pager duty

Cloud Platform

AWS, EC2, Auto Scaling, ALB, Lambda, CloudWatch, S3, Cloud Front, Route 53, RDS, SNS, SQS etc., Cloud Formation

AWS to Azure Migration

A large logistic data company took a strategic decision to migrate from AWS to Azure

• The applications were originally deployed on Elastic Beanstalk and the the migration involved analyzing and documenting 2 dozen applications and their elastic beanstalk dependencies
• Since the target platform was IaaS, rather than PaaS, the project involved refactoring the deployment approaches
• The migration exercise was used as an opportunity to streamline application deployment using better CI/CD processes
• Since a large number of applications were involved, infrastructure management was automated using Terraform and configuration management was done using Ansible

Cloud Platform

AWS, Azure (AWS Elastic Beanstalk, Azure VMs, VM scale sets, Azure Load Balancer, Azure Functions, Azure Resource Manager, Azure Monitor, Azure DevOps

Performance Optimization Consulting

A digital marketing agency was experiencing bottle necks in performance in certain parts of the application and vertical scaling measures did not yield the desired long-term results

• DB was identified as the bottle neck and mismatch of connection management between the application server and the DB was the key bottle neck
• Slow queries were analyzed and indexes were enhanced based on the current
• As a result of these improvements, the agency could scale down the DB server vertically bringing in cost optimization

Cloud Platform

AWS, EC2, CloudWatch

Other Technologies Used

Java, Apache Tomcat, PostgreSQL

Disposable Lab Environments on AWS

Designed and developed an API backend and Web UI for creating and managing on-demand lab environments to support various technology courses. The design included:

• Scheduled creation of lab environments for scheduled courses
• On demand creation of lab environments for individuals
• Life Cycle Management of lab environments

Platform

AWS Boto3, Apache, SNS, Lambda, CloudWatch, S3

Tools

HTML, CSS, JS, Jquery, Bootstrap, Python / Django, PostgreSQL