Skip to content

Volumes

The Volumes tab manages block storage — boot disks and data disks attached to your VMs. Two performance classes: NVMe HCI (default, hyper-converged) and Provisioned IOPS (NetApp-backed, with guaranteed IOPS).

Overview

Volumes list The Volumes tab — each row is one block volume with its size, type, attachment, status, and region.

Columns:

Column What it shows
Name Your name for the volume
Size GiB
Type NVMe HCI (default) / Provisioned IOPS
Attached to The VM name + mount path, or Unattached
Status In-use, Available, Creating, Error, Deleting
Created Relative timestamp
Region BD-DHK-1 / BD-CTG-1 / BD-SYL-1

Administration

Quotas

Project Settings → Quota:

  • Block storage GiB — total provisioned size cap
  • Volume count — number of volumes
  • Provisioned IOPS budget — total IOPS across all Provisioned-IOPS volumes (Provisioned IOPS uses a NetApp HA pair; per-project shares are tracked separately from generic GiB)
  • Snapshots per volume — default 100 rolling

Defaults

  • Default volume type for new volumes (per project) — useful to force Provisioned IOPS on databases-project but allow NVMe HCI for sandbox-project
  • Default encryption is always on (AES-256, cluster-managed keys); BYOK on the roadmap via OpenBao

Multi-attach (Provisioned IOPS only)

For clustered workloads that need a shared block target (rare — most cluster patterns use File Storage instead), Provisioned IOPS supports multi-attach: one volume mounted read-write to up to 8 VMs concurrently. Enabled at create time; cannot be turned on after.

Operation

Creating a volume

+ Create Volume. Fields:

  1. Name
  2. Size (GiB) — 1 to 16,384 (Provisioned IOPS: same range)
  3. Type — NVMe HCI / Provisioned IOPS
  4. Provisioned IOPS (only for Provisioned-IOPS type) — the IOPS target you want guaranteed
  5. Region
  6. Source — empty (default), restore from snapshot, or clone from another volume
  7. Tags

The volume enters Creating, then Available within seconds (longer for very large sizes).

Attaching

From the volume's detail panel, click Attach to server, choose the target VM and the device path (/dev/vdb, /dev/vdc, ...). For NVMe HCI volumes, virtio-blk; for Provisioned IOPS, virtio-scsi or NVMe-TCP depending on flavor.

Inside the VM, you'll then need to:

```bash

Linux: format and mount

sudo mkfs.ext4 /dev/vdb sudo mkdir /mnt/data && sudo mount /dev/vdb /mnt/data

Add to /etc/fstab with UUID for persistence across reboots

```

Resizing

Edit → Size → new GiB. Hot-grow is supported — the VM sees the new size immediately. You still need to grow the filesystem inside the VM (resize2fs, xfs_growfs, etc.). Shrink is not supported — to reduce size, create a smaller volume and migrate the data.

Snapshotting

Snapshot → New snapshot. Snapshots are redirect-on-write — instant creation, no IO penalty until divergence. They become images that you can clone into new volumes or images. See Images.

Detaching and deleting

Detach unmounts the volume from the VM (and the VM sees the device disappear). The volume goes to Available and can be attached elsewhere.

Delete is irreversible after a 7-day grace window. Snapshots survive deletion of the source volume unless explicitly deleted too.

Troubleshooting

Symptom Likely cause Fix
Cannot attach: "device path in use" The device path you picked is already mounted by another volume Use a different /dev/vdX path — list current mounts in the VM with lsblk
Attach succeeds but lsblk shows nothing virtio kernel module missing in the image Confirm your image was built with virtio-blk and virtio-scsi support
Volume creation stuck in Creating for > 10 min Underlying NetApp / HCI capacity contention Check Status; pick a different region; open a ticket for capacity
Latency higher than expected on Provisioned IOPS NetApp HA pair at capacity (rare) Open a ticket with the volume ID and timestamps; we'll measure and rebalance
Hot-grow worked, OS still shows old size Need to grow the filesystem inside the VM resize2fs /dev/vdb for ext4, xfs_growfs /mnt/data for xfs, NTFS expansion via Disk Manager on Windows
Detach fails: "volume in use" Process inside the VM has files open on the mount lsof /mnt/data → kill or quit; unmount; retry detach
Cannot delete: "has snapshots" Snapshots reference the volume Delete the snapshots first (or detach the lineage by promoting to image)

Pricing model

Per GiB-month + (for Provisioned IOPS) per provisioned-IOPS-month. See Pricing model. Snapshot storage is billed at Object Archive rates.