AWS eks documentation change
Summary
Enhanced documentation for zonal shift feature: Added EKS Auto Mode integration details, availability warnings, and clarified Karpenter support.
Security assessment
Changes focus on availability/resilience operations during AZ failures without security implications. No vulnerabilities or security features are documented.
Diff
diff --git a/eks/latest/userguide/zone-shift.md b/eks/latest/userguide/zone-shift.md index aad292667..f489c2210 100644 --- a//eks/latest/userguide/zone-shift.md +++ b//eks/latest/userguide/zone-shift.md @@ -43,0 +44,2 @@ After your Amazon EKS cluster has zonal shift enabled with ARC, you can start a + * If you’re using [EKS Auto Mode](./automode.html), EKS Auto Mode automatically stops provisioning new nodes in the impaired AZ. Voluntary disruption actions such as consolidation and drift that would affect the impaired AZ are also suspended. Pods with strict scheduling requirements that target the impaired AZ (such as persistent volume bindings, node affinity, or strict topology spread constraints) do not result in new node launch attempts to that AZ. + @@ -60,0 +63,4 @@ The following diagrams provide a high level overview of how EKS zonal shift ensu +###### Important + +A zonal shift shifts all in-cluster traffic away from an Availability Zone. If your workloads are not spread across multiple AZs with sufficient replicas, starting a zonal shift can itself cause an availability impact to your applications because there are no healthy endpoints to receive traffic. You should validate that your cluster environment can operate with one fewer AZ before enabling zonal autoshift or starting a manual zonal shift. + @@ -230 +236 @@ By using ARC zonal shift or zonal autoshift in your EKS cluster, you can better -EKS integrates with ARC, which provides the primary interface for you to accomplish recovery operations in AWS. To ensure that in-cluster traffic is appropriately routed away from an impaired AZ, EKS makes modifications to the list of network endpoints for Pods running in the Kubernetes data plane. If you’re using Elastic Load Balancing to route external traffic into the cluster, you can register your load balancers with ARC and start a zonal shift on them to prevent traffic from flowing into the degraded AZ. Zonal shift also works with Amazon EC2 Auto Scaling groups that are created by EKS managed node groups. To prevent an impaired AZ from being used for new Kubernetes Pods or node launches, EKS removes the impaired AZ from the Auto Scaling groups. +EKS integrates with ARC, which provides the primary interface for you to accomplish recovery operations in AWS. To ensure that in-cluster traffic is appropriately routed away from an impaired AZ, EKS makes modifications to the list of network endpoints for Pods running in the Kubernetes data plane. If you’re using Elastic Load Balancing to route external traffic into the cluster, you can register your load balancers with ARC and start a zonal shift on them to prevent traffic from flowing into the degraded AZ. If you’re using EKS Auto Mode, EKS Auto Mode automatically restricts node provisioning to healthy AZs. Zonal shift also works with Amazon EC2 Auto Scaling groups that are created by EKS managed node groups. To prevent an impaired AZ from being used for new Kubernetes Pods or node launches, EKS removes the impaired AZ from the Auto Scaling groups. @@ -243,0 +250,2 @@ If you are not pre-scaled and rely on provisioning additional nodes or Pods duri +If you are using EKS Auto Mode, EKS Auto Mode automatically provisions new nodes in healthy AZs to meet unschedulable Pod demand. However, new nodes still take time to launch and become ready. For the fastest recovery, we recommend pre-scaling your workloads across multiple AZs. + @@ -250 +258,5 @@ If you are running a stateful application, you must assess its fault tolerance, -**Does this feature work with Karpenter?** +**Does this feature work with EKS Auto Mode?** + +Yes. When you enable zonal shift on an EKS Auto Mode cluster, EKS Auto Mode automatically responds to zonal shift events. During a zonal shift, EKS Auto Mode stops provisioning new nodes in the impaired AZ, suspends voluntary disruption actions (such as consolidation and drift) that would affect the impaired AZ, and avoids launching capacity for Pods with strict scheduling requirements that target the impaired AZ. You don’t need any additional configuration beyond enabling zonal shift on the cluster. + +**Does this feature work with self-managed Karpenter?** @@ -252 +264 @@ If you are running a stateful application, you must assess its fault tolerance, -Karpenter support is available with ARC zonal shift and zonal autoshift in EKS with [Karpenter version 1.12 or greater](https://github.com/aws/karpenter-provider-aws). +Self-managed Karpenter support is available with ARC zonal shift and zonal autoshift in EKS with [Karpenter version 1.12 or greater](https://github.com/aws/karpenter-provider-aws).