Hybrid EKS Migration: Bridging On-Premises Infrastructure with AWS Cloud
Protagona helped a leading genomics company modernize its infrastructure in six weeks, delivering a hybrid Amazon EKS cluster that connects on-premises servers with AWS, enabling cloud-scale elasticity without abandoning existing hardware investments.
Industry
Healthcare
Teams & Services
DevOps
Tech & Tools
Hybrid Amazon Elastic Kubernetes Service (EKS), Oracle PCA On-Prem Hardware, Custom S3 Upload Application, Multiple VIF/ Direct Connect Connections
Key Data Points
The Vision
To build a secure, scalable hybrid cloud foundation that bridges existing on-premises infrastructure with the power of AWS, allowing a high-growth genomics company to handle massive data workloads without costly over-provisioning or rearchitecting what already works.
The Goal
Deliver a production-ready hybrid Kubernetes environment that unifies on-premises hardware with AWS managed services, standardizes container orchestration, and enables automated cloud bursting for high-volume genomics pipelines, all within a six-week engagement and without full cloud migration.
The Challenge
With a significant capital investment already in place on Oracle PCA X10 on-premises servers, a full cloud migration was simply not a practical option. The priority was finding a path to modernization that protected existing hardware ROI rather than rendering it obsolete.
At the same time, genomics pipelines demand massive, temporary bursts of compute power. Processing up to 90 TB of data per day requires scaling capacity that exceeds local hardware limits, yet permanently provisioning that capacity in the cloud would result in significant, unnecessary cost.
Compounding both challenges was the absence of a unified container orchestration platform. Without standardized tooling, deployments relied on manual processes and fragmented systems that slowed the team down and made consistent, efficient operations difficult to achieve.
The Solution
The project utilized AWS EKS Hybrid Nodes to bridge on-premises hardware with the cloud. This approach allowed the leading genomics company to keep its Oracle PCA X10 servers as worker nodes while managing them through an AWS-hosted control plane.
.png)