NetApp ONTAP 9: Failed-Disk Replacement and Ownership Checks

This source guide concerns ONTAP 9 cluster mode. Legacy Data ONTAP 7-Mode and early clustered 8.x commands differ; do not mix them with this procedure. Connect through cluster/node management SSH, or the appropriate SP/BMC console, using the platform’s supported access method. The source notes that large serial-console output may be slow and wrap long commands.

1. Confirm version and failed-disk identity

version -v
storage disk show -broken
system node run -node <NODE_NAME> -command sysconfig -a

Use the command output, serial number and shelf/bay labels to identify the exact failed disk. sysconfig is node-scoped and is supplementary hardware context, not a universal substitute for the platform’s physical replacement guide.

2. Check ownership policy

storage disk option show

The archived screenshot shows Auto Assign off on both nodes; firmware update and auto copy are on, and the policy is default. Its original CSDN author watermark includes a Chinese proper-name attribution alongside StorageExpert; that attribution is preserved in the otherwise English terminal image.

If automatic assignment is disabled, inspect unassigned disks and assign only the verified replacement to the correct node as required:

storage disk show -container-type unassigned
storage disk assign -disk <DISK_ID> -owner <NODE_NAME>
# Source example only:
storage disk assign -disk 0c.01.2 -owner node1

The source also lists bulk assignment and enabling autoassignment:

storage disk assign -all -owner <NODE_NAME>
storage disk option modify -node <NODE_NAME> -autoassign true

Bulk assignment can claim more than the intended replacement; do not use it as the default in a shared/partitioned environment. Autoassignment is a platform-dependent policy, not simply a universal preference for all production systems.

3. Locate and replace the physical disk

Reconfirm failure with storage disk show -broken. The source’s locate example uses -state on, but the checked ONTAP command reference documents -action on:

storage disk set-led -disk <DISK_ID> -action on

Follow the service procedure for the exact shelf/platform and confirm the physical disk before removal. After insertion, inspect whether the new disk is recognized and already assigned; a correctly autoassigned disk need not remain in the unassigned list. Assign ownership only if needed.

4. Verify spare or reconstruction state

storage aggregate show
storage disk show -container-type spare
storage aggregate show-status -aggregate <AGGREGATE_NAME>

An assigned disk that is not consumed by a reconstruction may become a spare. Verify suitability, ownership, pool/partition layout and readiness. ONTAP 9.5 and later support fast zeroing for applicable configurations; older behavior can take longer. Use the matching guide to confirm spare readiness rather than blindly zeroing a disk in use.

Replacement-drive qualifications

The source asks suppliers to remove stale ownership from replacement drives and prepare them before a site visit. This must not be confused with the statement “spares have no owner”: a usable spare can and normally does have an owning node, while unassigned is a separate state. Removing ownership in a running system has restrictions and requires the matching supported procedure; it is not impossible in every production scenario. Do not remove ownership from active data disks.

MetroCluster and Advanced Drive Partitioning require their own model/version-specific procedures. Check reconstruction progress and overall aggregate/HA health before completing the maintenance.

Adapted and translated from StorageExpert’s original CSDN article, identified by the source as licensed CC BY-SA 4.0. The English translation adds command, ownership and scope corrections. The original screenshot and author attribution are retained.