Performance Engg for an Apparel Giant

Performance engineering for design processing system for a large apparel manufacturer

Project Synopsis

A large, US based apparel company had an existing system which was used to process designs created by it's team of designers and update in the central repository. At any given time, there were multiple designers generating designs which needed to be processed in a fast manner. The existing system took fifteen minutes on average to pick up a newly created design. As a result, there was a big backlog of designs to be processed for the most part of the day. The system also blocked some designers or overwrote their designs while processing them.

The client was struggling with the following challenges:

  • Extremely slow performing system
  • Compromised productivity of designers because of the slow response time
  • Inaccurate processing of data currently being processed by designers

Rare Mile Solution

Rare Mile redesigned the design processing system to provide better performance. The heart of the system was a multi threaded engine which could create parallel processing tasks for faster assimilation of designs. The reference data of the system was cached effectively to reduce I/O time. Different components of the new system were designed in master/slave mode to aid horizontal scaling of the application for future requirements.

Rare Mile redesigned the application for high performance by employing the following means:

  • Designed different components of the system so that they could hosted on separate infrastructure and communicate with each other
  • Used multi threaded solution to better utilize the available hardware resources
  • Designed new means of identifying design work in progress to avoid inaccuracy in processing
  • Used Caching effectively to improve system performance

Project Highlights

  • Brought down the design processing time from 15 minutes to 15 seconds
  • Fixed price, milestone based assignment
  • Delivered a customized thread pool for super efficient use of hardware resources

About The Project

Performance driven redesign of application

Technologies Used

Microsoft .NET
Oracle Database
Custom Designed Thread Pool

Client Details