Scaling Guidelines for Kubernetes Access Node

For high availability and improved concurrent throughput of backups and other operations, instead of increasing the hardware specifications of your access node, add more access nodes (with identical hardware specifications) to your Kubernetes cluster. You can also add multiple access nodes to a server group, and then assign that server group to your Kubernetes cluster or application group.

The read speed from disks, the network throughput from worker nodes, and the network throughput on your access nodes affect the amount of data that is transferred in the backup window.

In managed cloud infrastructures, the underlying CPU and network credits that are granted to control plane nodes limit the amount of data that can be transferred concurrently. If you are reaching cloud-based resource quotas, consider scaling the control plane with additional nodes to support required backup throughput.