RBD: Snapshots >450 cannot be taken on a PVC in certain cases

[Rakshith-R]

Describe the bug

CephCSI follows the below logic before a k8s snapshot or pvc-pvc clone is created.

ceph-csi/internal/rbd/controllerserver.go

Lines 532 to 539 in a57fe08

	// check snapshots on the rbd image, as we have limit from krbd that an image
	// cannot have more than 510 snapshot at a given point of time. If the
	// snapshots are more than the `maxSnapshotsOnImage` Add a task to flatten all
	// the temporary cloned images and return ABORT error message. If the snapshots
	// are more than the `minSnapshotOnImage` Add a task to flatten all the
	// temporary cloned images.
	func flattenTemporaryClonedImages(ctx context.Context, rbdVol *rbdVolume, cr *util.Credentials) error {
	snaps, err := rbdVol.listSnapshots()

But if the cloned image[1] is in trash[2],
CephCSI still tries to add a task to flatten that cloned image (without checking if it is in trash) and fails.
The failure is only logged. CephCSI will not be able to create any more snapshots or pvc-pvc clones once the number hits 450+.

[1] cloned image: underlying a k8s snapshot or pvc-pvc clone)
[2] image in trash: this occurs when the image has child images of its own still alive.

Environment details

• Image/version of Ceph CSI driver : all
• Helm chart version : all
• Kernel version : all
• Mounter used for mounting PVC (for cephFS its fuse or kernel. for rbd its
  krbd or rbd-nbd) : rbd
• Kubernetes cluster version : all
• Ceph cluster version : all

Steps to reproduce

Steps to reproduce the behavior:

1. Create a rbd PVC.
2. Repeat both the belove steps greater than 450 times.
   • Create Snapshot.
   • Restore Snapshot into PVC.
   • Delete Snapshot.
3. No more snapshots above 450 can be created.

Actual results

The final snapshot creation fails.

Expected behavior

Snapshots should be created without error.

---

Propose solutions 1:

During k8s Snapshot Deletion :

• Check snapshot count on parent rbd image(underlying the parent PVC) [ Maybe skip this step and always flatten ? ]
• Add a task to flatten on rbd image underlying the k8s snapshot itself
• Move k8s Snapshot rbd image to trash and add a task to remove it [current process]

@idryomov @Madhu-1 @nixpanic Can you please provide your opinions/suggestions regarding this?

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed in a week if no further activity occurs. Thank you for your contributions.

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed in a week if no further activity occurs. Thank you for your contributions.

[github-actions]

This issue has been automatically closed due to inactivity. Please re-open if this still requires investigation.

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed in a week if no further activity occurs. Thank you for your contributions.

[github-actions]

This issue has been automatically closed due to inactivity. Please re-open if this still requires investigation.