LV Digital Migration Project (On-premise to AWS migration)
.jpg)
Key Challenges
LV Digital faced challenges with its legacy on-premises infrastructure, including multiple virtual machines, limited infrastructure flexibility, unclear server utilization, and the need to securely migrate terabytes of data and databases to AWS with minimal disruption.
Key Results
The migration to AWS enabled LV Digital to achieve a highly available, scalable, and secure cloud environment. With Amazon EKS, RDS, S3, automated scaling, centralized monitoring, and infrastructure-as-code, the organization improved operational efficiency, resiliency, security, and overall application performance.
Overview
LV Digital initiated a major modernization project to migrate its on-premises infrastructure to asecure, scalable, and highly available cloud-based environment on AWS (Frankfurt Region). Theproject, supported by Ankercloud, involved shifting legacy virtual machines and data services to amodernized architecture centered on Amazon EKS for containerized workloads, Amazon RDS fordatabase management, and Amazon S3 for object storage.
Challenges
- Legacy Infrastructure: Operations relied on an on-premises infrastructure consisting of multiple virtual machines for application servers, Apache Solr, Zookeeper, and management servers.
- Data Volume: A significant volume of data (Terabytes) required migration along with the database.
- Operational Limitations: The existing setup suffered from unclear server utilization metrics, and limited flexibility in infrastructure management.
Solution
- Cloud Architecture: Deployed a highly available architecture across Availability Zones (AZs) using Amazon VPC, Amazon EKS for containerized workloads, and Amazon RDS for MariaDB. Amazon RDS was deployed with Multi-AZ high availability, automated backups, encryption at rest, and defined backup retention.
- New HPA Setup: Kubernetes HPA automatically adjusts the number of pod replicas based on CPU and memory utilization. Node-level scaling is handled separately through the configured EKS node autoscaling mechanism.
- Neu ACM/SSL: ACM certificates were used to enable HTTPS on the CloudFront and/or Application Load Balancer endpoints, providing encryption in transit and domain authentication.
- Migration Tools: Utilized AWS DataSync for secure object storage migration to Amazon S3 and AWS Database Migration Service (DMS) for the database migration (Full load + CDC). Migrated data securely between the on-premises environment and AWS over a Site-to-Site VPN connection
- Automation: Shifted from manual "click-ops" to infrastructure-as-code using AWS CloudFormation for repeatable, versioned deployments.
- Operational Management: Implemented External Secrets Operator (ESO) for secure secret management via AWS Secrets Manager and integrated Auto Scaling to dynamically manage compute capacity based on workload demand.
Business Outcome
- Improved Scalability: The transition to EKS with auto-scaling policies ensures dynamic resource allocation, meeting changing workloads efficiently.
- Enhanced Monitoring: Established a comprehensive centralized monitoring and observability framework using Amazon CloudWatch (Logs, Metrics, and Alarms) for providing full visibility into the health and performance of resources.
- Operational Excellence: The shift to automated deployments and managed services minimized operational overhead, while modern security practices (e.g., SSM Session Manager for instance access) significantly improved the security posture.
- Resiliency: High availability across two AZs supports disaster recovery and business continuity.
- Content Delivery: Implemented Amazon CloudFront as the CDN layer across multiple distributions (including dedicated distributions for media and static file assets), with path-based cache behaviors and Lambda@Edge for bot protection (Datadome) on high-traffic listing/detail pages — improving global latency and offloading traffic from origin ALBs/EKS.
- SSL certificate: Provisioned SSL/TLS certificates in AWS Certificate Manager (ACM) to enable HTTPS, encrypt data in transit, authenticate the application domain, and provide secure client-to-ALB communication.
- HPA Setup: HPA automatically scales the number of pod replicas for a Deployment based on observed CPU/memory utilization (or container-level resource usage).
