About Customer
A b2b travel tech Giant headquartered in Delhi with a network of 50,000 travel agents and suppliers in over 100 countries. Its a hub for enabling the connectivity of travel agents with suppliers of travel products and services.
Problem Statement
-
Limited ability to scale due to the running of microservices on Amazon EC2 instances with Docker Swarm.
-
Faced inconsistency in platform performance.
-
Application setup was having issues with security and non-adherence to compliance standards.
-
Manual processes caused delay in deployments which resulted in longer time-to-market and higher chances of errors.
Solution Offered!
- Managed and provisioned AWS VPC and EKS using Terraform to enable a scalable infrastructure that ensures smooth performance during peak travel times.
- Used shared libraries to streamline development by reusing code across multiple projects, leading to quicker development cycles.
- Created a new IAM role for EKS via AWS IAM, to improve the overall security posture by managing the permissions and access controls.
- To create a secure and scalable infrastructure, container images were pushed to AWS ECR, facilitating the migration of microservices to AWS EKS.
- To proactively fix any potential issue that may hamper user experience, designed a custom monitoring solution with Prometheus, Alertmanager, and Grafana for the EKS cluster.
- Used helm charts to deploy the monitoring solution and applications over EKS to ensure consistent deployments and configuration for preventing costly downtimes.
- Leveraged Jenkins for the automation of deployments in EKS, and implemented AWS Backup Service for data loss prevention.
- To minimize latency and deliver faster load time, AWS Elastic Load Balancer (ELB) was implemented to ensure the high availability of microservices.
- For improved platform performance and cost optimization, AWS S3 was used for storing and managing container images, application artefacts etc.
- Implemented AWS RDS for a reliable and scalable managed database solution for ensuring smooth performance even at sudden traffic surges.
- Installed Karpenter as an autoscaling solution for Kubernetes to optimize resource utilization, consequently reducing the overall platform operational cost.
Final Outcomes
-
Completed the entire platform modernization setup in under 6 weeks.
-
The platform delivered consistent performance even during periods of traffic surges.
-
A 10x velocity improvement in time to market.
-
Reduced overall platform operational cost by 2x.
-
Centralized visibility of the application was achieved with a custom monitoring solution.
-
Enhanced the overall security, compliance and observability of the platform.
