Ubuntu to host containers in Amazon’s (EKS) for container portability

This article was last updated 5 years ago.


The benefits of Ubuntu optimisation by Amazon and Canonical now extend to Amazon’s Elastic Container Service for Kubernetes (EKS) with Ubuntu worker nodes hosting high performance Docker containers in AWS. This creates perfect portability for enterprise container workloads developed on Ubuntu workstations and operated on private infrastructure with Canonical’s distribution of upstream Kubernetes.

Amazon EKS is a fully automated Kubernetes cluster service on AWS. CDK is Canonical’s distribution of upstream Kubernetes that automates K8s deployment and operations for multiple infrastructure environments such as bare metal, VMware and OpenStack.

Ubuntu EKS worker nodes are built with the new ‘minimal Ubuntu’ base image, which dramatically shrinks the image size and security cross-section for Ubuntu in AWS.

Additional optimisations of this version of Ubuntu on AWS include:

  • Up to 30% faster kernel boot speeds
  • Elastic Network Adapter (ENA), supporting up to 25 Gbps network interfaces
  • Continuous image maintenance and security updates address critical issues
  • Improved i3 instance class support with NVMe storage disks for extreme IO
  • i3.metal support

Instructions for launching Ubuntu EKS clusters are available here.

Hybrid cloud operations with Kubernetes

Ubuntu is the most widely used container host OS – especially for deployments of upstream Kubernetes. With Ubuntu as their EKS container host, users have exactly the same runtime dynamics on AWS as on other Ubuntu-based Kubernetes deployments in private infrastructure. The Canonical Distribution of Kubernetes (CDK) on VMware, bare metal or OpenStack has exactly the same Canonical-supported kernels, libraries, tools and container runtime binaries. This greatly improves the efficiency of CI/CD and test pipelines which span workstations, private and public cloud infrastructure.

On AWS, the combination of Ubuntu, Amazon CloudFormation and Canonical’s cloud-init provide the flexibility to customise the EKS container hosts nodes to match enterprise requirements or integrate with enterprise infrastructure, such as custom security, management, monitoring and performance optimisations. Even though EKS is fully automated, the container host VMs can be customised with standard practices familiar to the majority of AWS users.

Enterprise users of Ubuntu container hosts in EKS and CDK on prem can expect:

  • Binary compatibility across the span of development, test and production
  • Greater productivity with fewer unexpected deltas in CI/CD pipelines
  • Greater velocity with the world’s most popular development tools and stack
  • Better security with a minimal attack surface
  • Improved economics and agility across private and public clouds

Ubuntu container hosts in EKS are supported by Canonical through the Ubuntu Advantage program, available through the AWS Marketplace. EKS users gain with the assurance they will receive the support coverage tailored to their specific environment and which can extend to both public and private deployments.

Ubuntu cloud

Ubuntu offers all the training, software infrastructure, tools, services and support you need for your public and private clouds.

Newsletter signup

Get the latest Ubuntu news and updates in your inbox.

By submitting this form, I confirm that I have read and agree to Canonical's Privacy Policy.

Related posts

Canonical at AWS re:Invent – What you need to know!

Though the Las Vegas Grand Prix has come to a close, the Canonical team is gearing up for the next big race at AWS re:Invent, slated for November 27-December...

Understanding AWS pricing

Have you ever wondered how AWS pricing actually works? You launch a startup or a new project in your organisation. You decide to use Amazon Web Services (AWS)...

Generative AI with Ubuntu on AWS. Part II: Text generation

In our previous post, we discussed how to generate Images using Stable Diffusion on AWS. In this post, we will guide you through running LLMs for text...