HomePortfolioOptimizing Cloud Resources and Ensuring Scalability with Amazon EKS

Optimizing Cloud Resources and Ensuring Scalability with Amazon EKS

AWS
Kubernetes
Blockchain
Finance

A developer of an Ethereum-based blockchain network turned to Altoros to optimize cloud resource utilization and scale elastically, while achieving zero-downtime upgrades.

Optimizing Cloud Resources and Ensuring Scalability with Amazon EKS

About the project

Brief results of the collaboration:

  • Delivered in just two weeks, an AWS-based prototype demonstrated how to reduce costs in cloud resource utilization by 70%.
  • With Amazon EKS, the company now has a roadmap for achieving ease of infrastructure management and operations automation.
  • The prototype also provided the means for enabling the platform to remain highly available under heavy workloads.
  • In addition, the customer fully automated the rollout of changes to new instances, thus reducing the time needed for upgrades from days to minutes.

The customer

The company is a creator of POA Network—an open platform for developing smart contracts based on Ethereum, providing an interface to view, confirm, and inspect transactions. With the market capitalization of $3.18 million (Feb 2020), POA traded 70% of their tokens worth $12.64 million during the initial coin offering (ICO) held in 2017.

The need

Initially, deployed to virtual machines, not all of the cloud resources were utilized efficiently, with most of them being idle. With dozens of thousands of users connected simultaneously, the platform did not scale elastically under heavy workloads. As a result, the company wanted to optimize its cloud resources and automate infrastructure management.

On top of that, the existing continuous integration / delivery pipeline did not provide automatic updates to all of the instances deployed, with manual efforts needed.

Therefore, the customer turned to Altoros, an experienced Kubernetes-certified service provider, to ensure its platform stays highly available in production during upgrades, while processing heavy workloads.

The challenge

POA Network is an innovative fintech startup growing constantly, with heavy workloads that need to scale elastically. Security, availability, and performance are important factors for them.

The solution

Evaluating the customer’s requirements and workloads, engineers at Altoros proposed to employ Amazon Elastic Kubernetes Service (EKS) to solve the existing bottlenecks. AWS provides a stable, secure, scalable, and reliable infrastructure with a broad range of managed services capable of addressing the customer’s challenges.

In the course of two weeks, our developers created a working prototype based on EKS, demonstrating how to assure a 70% reduction in the utilized cloud resources.

To improve availability during updates, Altoros suggested to containerize microservices that geared the platform. Using EKS helped to create isolated environments on each instance, ensuring the microservices seamlessly interact with each other without affecting the whole system.

After that, the team performed several load tests to measure the platform’s availability under increasing workloads.

By delivering a Bash script, our DevOps experts enabled the continuous integration / delivery pipeline to provide automatic updates across all of the hosted blockchain networks. This allowed for reducing deployment time from days to minutes.

The outcome

By employing vast expertise in infrastructure optimization, Altoros helped the customer to reduce cloud costs by 70%. Relying on a proven methodology for implementing production-grade Kubernetes clusters, the customer was able to ensure its platform remains highly available and secure, while scaling elastically. By automating the rollout of updates to all of the instances on the blockchain network, the company sped up the delivery of new functionality from days to minutes. With a broad range of managed services provided by AWS, the customer can go further in eliminating operational efforts, thus simplifying the deployment process and optimizing infrastructure costs.

Technology stack

Platforms

Amazon Web Services, Amazon Elastic Kubernetes Service

Programming language

Bash

Frameworks and tools

Amazon Elastic Container Registry, Ethereum Virtual Machine, Docker, CircleCI, Phoenix

Database

PostgreSQL

/
01

Want to develop something similar?

Preloader
Ryan Meharg

Ryan Meharg

Technical Director

ryan.m@altoros.com650 265-2266

4900 Hopyard Rd. Suite 100 Pleasanton, CA 94588