Cloud Init
AWS EKS Auto Mode: Key Highlights from re: Invent

AWS EKS Auto Mode: Key Highlights from re: Invent

New to Cloud? AWS EKS Auto Mode Simplifies Your Start.

Cloud Commander's photo
Cloud Commander
·

4 min read

Overview

As the AWS Re: Invent is happening this week, many new products and features will be announced. One of the initial announcements that surprised me was the general availability of AWS EKS Auto Mode. Which further abstracts the complexity of managing EKS clusters on AWS.

The EKS Auto Mode automates cluster management without requiring extensive Kubernetes knowledge since it selects appropriate compute instances, dynamically scales resources, continuously optimizes prices, controls core add-ons, patches operating systems, and connects with AWS security services. In EKS Auto Mode, AWS takes on more operational responsibility than the customer-managed infrastructure in your EKS clusters. In addition to the EKS control plane, AWS will configure, maintain, and protect the AWS infrastructure in EKS clusters required to execute your apps.

EKS Auto Mode enables the following Kubernetes capabilities in your EKS cluster:

  • Compute auto-scaling and management

  • Application load balancing management

  • Pod and service networking and network policies

  • Cluster DNS and GPU support

  • Block storage volume support

Now you can just get started with filling out 5-7 fields on the console

Here are my views on this service

  1. I welcome these features as they will greatly benefit startups onboarding themselves on eks, but they do not have much experience optimizing it for production-heavy workloads.

  2. Newcomers will also be encouraged to try out this and deploy their applications or workloads on EKS, as now they don’t have to learn all the terminologies and settings eks has just to get started. Now, it's just as simple as a few networking fields, and off you go!

  3. For seasoned professionals, it might not be that interesting of a prospect from a scaling point of view. Also, people today manage their clusters with different available cloud-native apps like Keda, Kyverno, Argo, etc. This gives them the flexibility to do things their way, which might be specific to the needs of the application they are working on.

  4. Also, the auto compute part, where AWS will manage all the underlying nodes and also give you the option to have some self-managed nodes as well, is interesting, as organizations might have applications that require specific nodes with specific support for a particular processor or OS architecture.

    For example, if you want to run something on a custom node you can do with this

    You can add this nodeAffinity to Deployments or other workloads to require Kubernetes not schedule them onto EKS Auto Mode nodes.

       affinity:
     nodeAffinity:
       requiredDuringSchedulingIgnoredDuringExecution:
         nodeSelectorTerms:
         - matchExpressions:
           - key: eks.amazonaws.com/compute-type
             operator: NotIn
             values:
             - auto

How Auto mode’s node management will come in comparison to karpenter is still something we have to test against our workloads.

Customizable NodePools and NodeClasses: If your workload requires changes to storage, compute, or networking configurations, you can create custom NodePools and NodeClasses using EKS Auto Mode. While default NodePools and NodeClasses can’t be edited, you can add new custom NodePools or NodeClasses alongside the default configurations to meet your specific requirements.

  1. But, for testing purposes, this auto mode is great as we now don’t have to worry about all the other 100 things before we can start testing some new integrations or add-ons.

  2. Also, this auto mode will be a great help for organizations that want to quickly spin up a well-optimized managed cluster for a short delivery cycle or need infrastructure to be ready for customer delivery without time to test what works best.

  3. It will also be interesting to see the cost optimization on this auto mode, as infra administrators sometimes struggle to reduce and optimize the cost aspects to that extent in a ploy to make things more secure and highly available.

Conclusion

This re: Invent week brings new features and resources to our plate; testing and experimenting with these will be fun. I will be posting a few more blogs on exciting topics coming my way from re: invent this week, so follow me to stay updated.

Streamline Kubernetes cluster management with new Amazon EKS Auto Mode | AWS News Blog

Create cluster with EKS Auto Mode - Amazon EKS

AWS re:Invent 2024 live news and session updates

EKSreInventAWSk8s