What is Storage Autoscaling?
Storage Autoscaling is an auto-scaling storage solution that dynamically manages Persistent Volumes (PVs) for Kubernetes workloads. It automatically adjusts the size of storage volumes based on actual usage, helping to reduce cloud storage costs and improve resource efficiency.
What Kubernetes environments does Storage Autoscaling support?
Storage Autoscaling supports Amazon EKS (Elastic Kubernetes Service).
Which workloads does Storage Autoscaling support ?
All workload types are supported. StatefulSet workloads are fully supported; others have beta support.
How do I install Storage Autoscaling in my Kubernetes cluster?
Storage Autoscaling is deployed using a Helm chart. For more information, see Deploy PVs Autoscaling. Storage Autoscaling will integrate with your existing Storage classes and begin managing them automatically. Existing PVCs will need to be migrated to work with Storage Autoscaling.
What kind of applications benefit most from Storage Autoscaling?
The applications that benefit most are those that require dynamic storage management, especially:
Databases such as MySQL, MongoDB, Prometheus
CI/CD pipelines that generate large amounts of temporary data such as Argo, Tekton, Jenkins
Big data or AI applications with fluctuating storage demands such as AWS EMR, Spark, Ray
How does Storage Autoscaling affect Kubernetes performance?
Storage Autoscaling is designed to minimize any performance impact. It operates in the background, making volume adjustments without affecting the application’s availability or performance. The resizing operations are triggered based on specific thresholds to avoid unnecessary resource consumption.
By serializing the volumes on the filesystems, Storage Autoscaling multiplies the number of volumes, thereby increasing overall performance.