Case Study

Know how we are thinking and executing

Smart Messaging - IDE

Executive Summary

A Hi-Tech messaging company wanted to setup IDE for users to write their own bots and test them online which they can integrate with Facebook, Telegram, WhatsApp etc.

The Challenge
  • Standalone application with multiple web servers on single system
  • Improper architecture with multiple SPOFs (single point of failure)
  • Deployment Automation missing
  • Application Logging visibility was missing.
  • Scaling was not handled properly.
Insight Action
  • Did the in-depth analysis of what application does what are the component of the application which are loosely and tightly coupled.
  • Worked with dev team to separate the component to make micro serviced.
  • Configured and Installed the Docker with Mesos as management server for all these application.
  • Identified the SPOFs and help them to create the HA Architecture to eliminate the SPOF at all the levels.
  • Implemented the Automated CICD solution to take care for all application deployment with Jenkins jobs using AMIs as the base
  • Implemented the Centralised logging Server for Engineering team so that no direct access to production system needed.
  • Identified and projected the all-round needs for increased growth and configured the AutoScaling to handle the scale.
  • Made use of Terraform for templatisation of environment.
The Impact
  • Container based applications helped to take any type of load without affecting the cost.
  • Planned growth within the company, the needs of which were anticipated and met.
  • CICD implementation reduced the turn around time for deployment
  • Centralised logging were appreciated during Audits.
  • Team was able to concentrate more on development than the infrastructure issues.
STACK Used

For success of project TECHPARTNER used below Services

  • Docker was used for the microservices architecture.
  • Ansible was used for the config managment (Patching, Security updates etc) of servers.
  • Git was used to keep track the code repository.
  • Jenkins was used with webhook constantly checking git changes and automatically execute test cycle & deploy the successful build in Dev/Staging/Production Environment.
  • HAProxy was used for Loadbalancing between the Dockers microservices
  • Messos was used for docker management
  • Terraform was used to templatise the infrastructure footprint.