Volumes in a Managed Kubernetes cluster

Last update: May 30 2026

Managed Kubernetes uses local and network disks from the cloud platform.

You can use local and network volumes in a Managed Kubernetes cluster as a boot disk and for data storage.

When choosing a disk, consider the specifics of using local and network volumes in Kubernetes.

Using network volumes and local disks in Kubernetes

	Local disks	Network volumes
Advantages	Low data access latency	Ability to store persistent data that needs to be preserved during pod restarts and migrations
Limitations	No data accessibility between pods	Additional network latency for data access
Objects in Kubernetes	Volume and VolumeMounts, PersistentVolumes and PersistentVolumeClaims (when installing the CSI driver manually)	PersistentVolumes and PersistentVolumeClaims
Usage	Boot volume Data storage	Boot volume Data storage

Boot volumes

Boot (system) volumes are used to start and initialize the operating system and applications in Kubernetes containers.

You can use local or network disks from the cloud platform as boot disks. You can select the disk size for a network volume, whereas the size for a local disk is fixed.

You select the boot disk when creating a cluster and cannot change it after the cluster is created. When choosing, consider that:

• network volumes of the Universal SSD v2 and Fast SSD v2 types do not support changing the IOPS limit in a fault-tolerant multi-zonal cluster;
• in some fixed configurations with GPUs, only a local disk is available.

When nodes are reinstalled, all data on the boot disks is deleted.

Nodes are reinstalled during:

• minor cluster version upgrades;
• manual node reinstallation;
• automatic patch version updates;
• automatic node recovery.

Data storage volumes

To store data in Managed Kubernetes clusters, we recommend using PersistentVolume (PV) on network volumes.

If you have installed the CSI driver yourself and created a PV on a local disk, the data will be deleted when the node is removed.