Happy 5th Birthday Amazon EKS!

Today we’re thrilled to celebrate the 5th anniversary of Amazon Elastic Kubernetes Service (Amazon EKS), and it’s an opportune moment to reflect on our journey so far. Since its launch in 2018, Amazon EKS has served tens of thousands of customers worldwide in running resilient, secure, and scalable container-based applications. Amazon EKS, using upstream Kubernetes, is the most reliable way to run Kubernetes in AWS, on-premises data centers, and at the edge.

Reflecting on five years of milestones

Launched five years ago on June 5th, Amazon EKS was developed primarily in response to customers’ feedback that managing Kubernetes is difficult. Rick Sostheim, Principal Engineer on the Amazon EKS team, said, “Given the reception for the EKS beta offering, we were extremely excited about the potential, but nothing really could prepare us for what was to come next. The journey has been like riding a rocket, and yet I still feel like it’s Day 1 and where we’re going can only be imagined!” These sentiments reflect the progress and enthusiasm that have characterized our commitment to Amazon EKS customers.

Over these past 5 years, Amazon EKS has seen a number of milestones, each contributing to its resilience, security, and robust set of capabilities. Initially, we focused on creating a highly available and resilient control plane, while expanding the set of features for customer workloads in the data plane:

  • Easy cluster updates: Amazon EKS provides managed, in-place cluster upgrades, which allows you to use the Amazon EKS API to update your clusters to the latest version.
  • Managed node groups and AWS Fargate: Customers have the flexibility to run their workloads on Amazon EKS with managed node groups powered by Amazon Elastic Compute Cloud (Amazon EC2) instances and auto-scaling groups (ASGs). They can also utilize AWS Fargate to allow AWS to handle the patching, scaling, and security of the underlying nodes and simply focus on their workloads.
  • Cluster auto scaling with Karpenter: With Karpenter, customers can improve application availability and cluster efficiency with dynamically provisioned compute, even Spot instances, in response to the needs of your application workloads. A key feature, workload consolidation, results in substantial cost savings for customers.

Our journey continued with an emphasis on cluster operations, with the aim of providing a fully managed experience across the entire cluster lifecycle. This is realized through essential features like Amazon EKS Add-ons to manage operational software like the vpc-cni, CoreDNS and more, and IAM Roles for Service Accounts (IRSA), which can be used to securely connect to services like Amazon VPC Lattice using AWS Identity and Access Management (AWS IAM). Lastly, open-source initiatives within AWS led to the development of Amazon Controllers for Kubernetes (ACK) and EKS Blueprints for Terraform, CDK, and Crossplane making it easier for our customers to use Amazon EKS to leverage the entire AWS ecosystem. Amazon EKS empowers customers to deploy, manage, and operate all the operational software and add-ons necessary for production-ready Kubernetes deployments.

Moreover, we continue to demonstrate our commitment to meeting customers where they are. This is highlighted by the launch of Amazon EKS Anywhere in 2021 and our support for Amazon EKS on AWS Outposts.

The road ahead

As we chart the course for the future, our vision remains steadfast — listen to and prioritize the needs of our customers, consistently optimize our service to pass the savings to our customers, and increase contributions to upstream efforts to ensure the viability and safety of open-source software (OSS). Notably, we are increasingly focusing on the onboarding experience for Kubernetes on AWS. We aim to further reduce barriers to entry and expand our services to more Regions, supporting our customers wherever they choose to run their workloads — be it on the cloud, private data centers, or at the edge. Our key focus areas include:

  • Enhancing the Amazon EKS user experience for customers at all stages of their Kubernetes journey, from day-one operations to advanced cluster management at scale. This goal is spearheaded by continual improvements to the core Amazon EKS API, CLI, and UI experience. We’re also expanding our focus on providing application-ready Amazon EKS clusters, with all the operational software needed to run production-grade clusters.
  • Sustaining our investment in upstream development efforts, recognizing that the success of our services is intertwined with the success of the underlying open-source projects. In the past year, AWS has made significant contributions to a range of projects, including containerd, Cortex, etcd, Fluentd, nerdctl, Notary, OpenTelemetry, Thanos, and Tinkerbell.
  • Assisting customers in optimizing their cloud spend with cost-effective solutions, from our recent partnership with Kubecost, developments on open-source Karpenter to enhance cost efficiency, and continual guidance offered via our Amazon EKS Best Practices Guide.
  • Streamlining the deployment and management of third-party software on Amazon EKS clusters with ongoing improvements to our Amazon EKS add-ons framework. The recent integration of Amazon EKS Add-ons with AWS Marketplace simplifies the process of subscribing to and deploying partner solutions.

We’re excited about the future of Amazon EKS and encourage you to track our upcoming features, provide feedback, and influence the direction of Amazon by participating directly on our public AWS Container Services Roadmap.

Learn more about Amazon EKS

As we commemorate five remarkable years of Amazon EKS, we extend our gratitude to our customers, partners, and contributors who have been instrumental in this journey. For those yet to explore the potential of Amazon EKS, we invite you to start your journey today.