Disk partitioning in Managed Kubernetes on dedicated servers

In a Managed Kubernetes cluster on dedicated servers, only one disk type is partitioned automatically — the one with the highest performance. Disk types, ordered from highest to lowest performance: SSD NVMe ⟶ SSD SATA ⟶ HDD SATA.

The disks in the cluster are combined into RAID1 — all data is simultaneously written to two physical disks in the RAID array. Both disks in the array are partitioned.

The following partitions are allocated by default: root partition /, /boot and /storage.

You can only change disk partitioning when creating a cluster. You can opt out of the /storage partition and change the size of the root / partition. To see the list of partitions, their default sizes, and value limits, refer to the table Disk partitions and their size.

In an existing cluster, you can view the disk partitioning in the control panel: in the top menu, click Products and select Managed Kubernetes → cluster page → Cluster composition tab → click the node name → Operating System tab → Partitioning.

Disk partitions and their size

	Default size	Minimum size	Maximum size
/	100 GB	30 GB	Calculated using the formula: disk size (in GB) - 1 GB for the /boot partition - 1 GB for the /storage
/boot	1 GB*	1 GB	1 GB
/storage	Calculated using the formula: disk size (in GB) - 1 GB for the /boot partition - 100 GB for the /storage	1 GB	Calculated using the formula: disk size (in GB) - size of the root / partition - 1 GB for the /boot

* The size of the /boot partition cannot be changed.