{{CANONICAL}}
← Back to Tech News

Karpenter now supports Amazon Application Recovery Controller zonal shift

Amazon Web Services has integrated its Application Recovery Controller (ARC) zonal shift capabilities with Karpenter, the open-source Kubernetes node provisioning tool for Amazon EKS clusters. The integration allows organizations to automatically redirect network traffic away from impaired Availability Zones while preventing Karpenter from provisioning new compute capacity in affected zones during outages or performance issues. The new functionality supports both manual zonal shifts, where administrators can manually trigger traffic redirection, and zonal autoshifts, which automatically detect AZ impairments and redirect traffic without human intervention. When a zonal shift is activated, Karpenter halts voluntary node disruptions like consolidation and drift in the impaired zone and prevents scheduling dependencies that would require the affected zone. The system also includes practice runs to verify cluster functionality with reduced AZ capacity before real incidents occur. To enable the feature, administrators need to set the ENABLE_ZONAL_SHIFT configuration in their Karpenter settings, with the integration leveraging existing EKS cluster ARC resources without requiring additional custom configurations. When zonal shifts expire or are manually canceled, Karpenter automatically resumes normal provisioning and scheduling operations across all available zones.

Why It Matters

This integration addresses a critical challenge in multi-zone Kubernetes deployments where infrastructure failures can cascade into application outages. By automating the coordination between traffic management and compute provisioning during AZ impairments, organizations can achieve more resilient EKS deployments with reduced manual intervention. The feature is particularly valuable for enterprises running mission-critical workloads that require high availability guarantees and automated disaster recovery capabilities.

Read Original Release →
Note

This summary is generated using AI analysis of the original press release. Always refer to the original source for complete details.