AWS Extended EKS Support: A Costly Band-Aid for Kubernetes Clusters
Amazon Web Services (AWS) recently announced extended support for Amazon Elastic Kubernetes Service (EKS) versions (starting April, 2024), allowing customers to use older versions of Kubernetes for an additional 12 months. While this may seem like a convenient option, it comes with a hefty price tag and several drawbacks that customers should carefully consider before opting for it.
The Costly Catch: Automatically charged. No Opt-Out
AWS charges $0.60 per cluster per hour for EKS versions in extended support, which is SIX times! the cost of running a cluster on a standard support version ($0.10 per cluster per hour). This can add up to significant expenses for customers with large EKS deployments.
Starting April 2024: Every 25 clusters will cost an extra $100K/year. Automatically charged without an option to opt out.
Dependency Hell & Compatibility Issues:
While a specific Kubernetes (K8s) version might still be supported under extended support, the rest of the cloud-native software that customers are using probably isn’t. This means that at some point, things will break due to compatibility issues and lack of support for out-of-stream services.
Extended support for EKS versions does not guarantee compatibility with all cloud-native software and services. As newer versions of Kubernetes and other cloud-native components are released, they may introduce breaking changes that are not supported in older versions. This can lead to dependency hell, where customers are forced to maintain multiple versions of software and services to ensure compatibility with their EKS cluster.
For example, a customer using an extended support version of EKS may encounter compatibility issues with newer versions of popular cloud-native tools, Kubernetes Operators, OpenTofu/Terraform providers or any other tool that runs on the Kubernetes platform. This can make it difficult to manage and update their EKS cluster, and can lead to downtime or data loss.
Missed Out on Innovation:
Extended support for EKS versions can disincentivize customers from upgrading to newer versions, which means missing out on performance optimizations or new features in k8s or other services that are leveraging k8s as part of their engine. This can have a negative impact on the security, innovation, efficiency, and overall user experience of EKS clusters.
Uptime Risks:
While AWS provides security patches for EKS versions in extended support, these versions may not receive the latest security enhancements and features available in newer EKS versions. This could increase the risk of security vulnerabilities and exploits, potentially leading to downtime and data breaches.
Use Cases to Avoid K8s Extended Support:
Cost-Sensitive Applications: Customers with cost-sensitive applications should avoid extended support due to the significant additional cost it incurs. Upgrading to newer EKS versions more frequently may be a more cost-effective option in the long run.
Security-Critical Applications: Applications that handle sensitive data or require high levels of security should not be run on EKS versions in extended support. The lack of the latest security enhancements and features for both k8s and the other tools in the ecosystem — could compromise the security of these applications.
Applications Requiring the Latest Features: Customers who rely on the latest Kubernetes features or leverage 3rd party tools for their applications should not use extended support. Extended support versions may not have access to the latest features and functionality, which could limit the capabilities of these applications.
Conclusion:
AWS extended support for EKS versions may seem like a convenient option, but it comes with a hefty price tag and several drawbacks. Customers should carefully evaluate their application requirements, security posture, and financial constraints before opting for extended support.
In many cases, upgrading to newer EKS versions more frequently is the best way to stay cost-effective and it is also the most beneficial option for maintaining secure, performant, and up-to-date tech stack.
Feedback
Are you a Head of DevOps or Platform Engineering struggling to keep up with the constant version changes of your stack?
If so, I understand your pain. The rapid pace of innovation in the cloud-native ecosystem can make it difficult to keep your Kubernetes clusters and ecosystem up to date and secure.
I’m here to help. Please don’t hesitate to reach out to me on any social media platform.
I’m always happy to discuss your challenges and provide guidance and help.
LinkedIn: https://www.linkedin.com/in/talkimhi/
Twitter: @talkimhi
Let’s work together to tame the chaos of version changes.
