Troubleshoot storage

Each section below describes one symptom, its cause, and how to resolve it. For the error codes the API returns, see Error codes.

Volume reports ERROR after create or import

Symptom: A new volume reports status: ERROR instead of AVAILABLE, with a statusReason of the form the operation failed (<code>).

Cause: The platform couldn’t provision the volume. There’s no usable disk behind it.

Resolution:

1. Delete the volume. Deleting an ERROR volume is allowed. See Delete a volume.
2. Create or import the volume again.

For import failures, see Import fails with ERR_IMPORT or ERR_IMPORT_IMAGE_TOO_LARGE.

Volume reports ERROR after delete

Symptom: A deleted volume stays in your volume list with status: ERROR instead of disappearing.

Cause: The platform couldn’t remove the volume’s disk. The disk may still exist, so the volume still counts against your quota.

Resolution:

1. Delete the volume again.
2. If it returns to ERROR again, contact support with the volume’s URN and its statusReason.

Resize reports FAILED

Symptom: After a resize, the volume reports modificationStatus: FAILED with a modificationStatusReason, and sizeInGib still shows the old size.

Cause: The platform accepted the resize but couldn’t apply it. The volume’s data is intact and it’s still usable at its old size.

Resolution:

1. Request the resize again. See Resize a volume.

Resize is refused

Symptom: resize-volume fails immediately with an error such as volumes can only grow, the volume is already <n> GiB, or a volume in state <state> cannot be resized.

Cause: The requested size isn’t larger than the current size, or the volume isn’t AVAILABLE or INUSE, or another resize is still in progress. Nothing about the volume changed.

Resolution:

1. Ask for a size larger than the current sizeInGib. Volumes can’t shrink.
2. Wait for a CREATING volume to become AVAILABLE, or for modificationStatus to leave RESIZING, then retry.

Volume delete is refused because it’s attached

Symptom: delete-volume fails with the volume cannot be deleted while an instance attachment still claims it: <attachment-id>.

Cause: The volume is attached to an instance and reports INUSE.

Resolution:

1. Find the instance. See Find the instance a volume is attached to.
2. Detach the volume. See Attach and detach volumes.
3. Delete the volume again.

Volume delete is refused because of snapshots

Symptom: delete-volume fails with the volume cannot be deleted while a snapshot of it still exists: <snapshot-id>, or the same message naming several snapshots.

Cause: A snapshot depends on its volume’s data. Every snapshot that isn’t DELETED blocks the delete, including one in ERROR.

Resolution:

1. Delete each named snapshot, unprotecting it first if needed. See Delete a snapshot.
2. Wait until each reports DELETED.
3. Delete the volume again.

Volume still exists after its instance was terminated

Symptom: After you terminate an instance, one of its volumes is still AVAILABLE and billed, and retentionNote is set.

Cause: The volume couldn’t be deleted with the instance, most often because a snapshot of it exists. The platform keeps the volume instead of deleting your snapshot.

Resolution:

1. Delete the volume’s snapshots, then the volume. See Delete a volume.

Snapshot can’t be used as a clone source

Symptom: Creating a volume from a snapshot, creating an image, or launching from an image fails with snapshot is not available or snapshot is not protected.

Cause: A clone source must report status: COMPLETED and protectionStatus: PROTECTED.

Resolution:

1. Wait for a PENDING snapshot to reach COMPLETED.
2. Protect the snapshot, or protect it again if protectionStatus is FAILED. See Create and protect a snapshot.

If the error says the size must be greater or equal than used snapshot size, ask for at least the snapshot’s volumeSizeInGib.

Snapshot delete is refused

Symptom: delete-snapshot fails with snapshot must be unprotected before it can be deleted.

Cause: The snapshot is PROTECTED, in a protection change, or its protectionStatus is FAILED.

Resolution:

1. Unprotect the snapshot and wait for UNPROTECTED. See Delete a snapshot.
2. Delete it again.

Import fails with ERR_IMPORT or ERR_IMPORT_IMAGE_TOO_LARGE

Symptom: An imported volume reports ERROR with statusReason the operation failed (ERR_IMPORT_IMAGE_TOO_LARGE) or the operation failed (ERR_IMPORT).

Cause: ERR_IMPORT_IMAGE_TOO_LARGE means the image’s virtual size is larger than sizeInGib. ERR_IMPORT means the download, conversion, or write failed: the source returned an HTTP error, sent nothing for 60 seconds, or the import took longer than 25 minutes.

Resolution:

1. Delete the failed volume.
2. Import again with sizeInGib at least the image’s virtual size, or from a faster or reachable source. See Import a disk image as a volume.

Request refused with PRODUCT_UNAVAILABLE

Symptom: create-volume or import-volume fails with PRODUCT_UNAVAILABLE and a message such as storage class 'st1' is not available - use one of: gp1.

Cause: The storage class is published but can’t be selected.

Resolution:

1. Use a class whose available is true, or omit the class to get gp1. See Storage classes.

Request refused with PAYMENT_METHOD_REQUIRED

Symptom: Creating, importing, or resizing a volume, or creating a snapshot, fails with PAYMENT_METHOD_REQUIRED.

Cause: The project’s billing account can’t pay. For card-managed billing, no verified payment method is on file. Nothing was created or changed.

Resolution:

1. Add a payment method. See Add a payment method.
2. Retry the request.

Request refused because a quota is exceeded

Symptom: A create, import, resize, or snapshot request fails with a message such as organisation quota exceeded: volume storage (GiB) usage is <n>, requesting <m> more would exceed the limit of <limit>.

Cause: The request would take your organization or project over its volume storage or snapshot quota. Volumes in CREATING, DELETING, and ERROR count too.

Resolution:

1. Delete volumes or snapshots you no longer need, including ones in ERROR.
2. Retry the request, or ask for a higher limit. See Limits.

Request refused because the organization isn’t active

Symptom: A create, import, resize, or snapshot request fails with organisation is <status>, not active.

Cause: Your organization hasn’t been approved yet, or is no longer active.

Resolution:

1. Check your organization’s status. See Review your organization’s status.