As the competition in the retail space grows relentlessly, every step to enhance operations contributes to staying ahead. Recognizing this, one of the world’s largest logistics companies decided to provide a competitive advantage to their retail customers by offering improved shipment visibility. This involved building a high-availability and high-performance solution that could provide more precise delivery predictions.
The rich data generated within the logistics network of the company contained powerful insights. The company decided to apply these insights to remove barriers in the supply chains of their retail customers. The goal was to shrink the supply chain considerably. This vision led the client to envisage a delivery prediction solution – a market differentiator for their retail customers.
Key Challenges
However, there were multiple challenges to be overcome:
- Despite moving to a new application programming interface (API) based architecture, certain performance issues impacted the onboarding of new retailers
- Longer downtime during deployments began to affect reliability and end-user experience
- The solution used polyglot programming on multiple Azure cloud platforms with infrastructure as a service (IaaS) and platform as a service (PaaS).
- Siloed and inconsistent DevOps implementations were leading to quality and security concerns
Ready to experience?
TALK TO EXPERTS
The Solution
Standardized DevOps pipelines with a focus on automated performance testing and innovative techniques deliver a high-performance solution with greater reliability
We enabled the client to enhance the performance and reliability of the solution through these steps:
- Lifecycle automation: The team implemented dynamically scaled DevOps pipelines to achieve continuous integration and continuous delivery for multiple technologies. The standardized pipelines included automated quality and security checks in addition to automated acceptance tests.
- Exhaustive performance testing: Performance test environments were provisioned using infrastructure as code (IaC) along with a custom synthetic test data generator.
- Ensuring high scalability of application: We defined replicas to ensure the availability of the Kubernetes pods and implemented horizontal pod auto-scaling (HPA) for variable loads
- Automated and secured zero-downtime deployment: Blue-green deployment strategy via continuous delivery pipelines decreased the downtime to zero
Benefits
The logistics leader engaged with Infosys to help achieve their vision. The Infosys team enabled them to provide their retail customers with a superior solution for demand prediction. The company achieved:
50% effort reduction for automated build and deployments
70% effort and time reduction in performance testing
10 thousand transactions per second (TPS) for APIs and 15 thousand TPS for backend services
API response time reduced to milliseconds
Zero downtime during deployments
Certain key measures implemented by the Infosys team enabled the client to enhance their solution manifold:
- Standardized DevOps pipeline significantly reduced the build time and deployment time while meeting the company’s quality and security goals
- The IaC approach made faster infrastructure provisioning and scaling possible
- The team achieved zero-downtime deployments by automating all the deployment steps and leveraging the blue-green deployment strategy
- We also enabled the ability to extensively test the performance of APIs under variable loads and scenarios with automated performance testing
- By using automated alerts and dashboards, we managed to reduce mean time to detect (MTTD). Improved logging and automation allowed faster time to recover from issues. This further improved the overall availability and reliability of the solution
