NetApp

ONTAP Aggregate Destroy and Recreate Runbook

Published May 25, 2026 | 10 min read

Destructive Change Warning

Deleting an aggregate is destructive. NetApp documents that storage aggregate delete deletes a storage aggregate and fails if volumes are present. It also notes that if an object store is attached, aggregate deletion deletes objects in the object store: storage aggregate delete.

Use this runbook only after all data has been migrated, protected, or explicitly approved for deletion.

Prechecks

curl -k -u admin:'<password>' \
  -X POST \
  "https://cluster.example.com/api/storage/aggregates" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "aggr_new01",
    "node": { "name": "node01" },
    "block_storage": { "primary": { "disk_count": 10, "raid_type": "raid_dp" } }
  }'

Stop if any user, root, DP, clone, or system volume still depends on the aggregate.

CLI Destroy Process

Confirm no volumes remain:

curl -k -u admin:'<password>' \
  -X POST \
  "https://cluster.example.com/api/storage/aggregates" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "aggr_new01",
    "node": { "name": "node01" },
    "block_storage": { "primary": { "disk_count": 10, "raid_type": "raid_dp" } }
  }'

Offline the aggregate if required by your procedure:

curl -k -u admin:'<password>' \
  -X POST \
  "https://cluster.example.com/api/storage/aggregates" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "aggr_new01",
    "node": { "name": "node01" },
    "block_storage": { "primary": { "disk_count": 10, "raid_type": "raid_dp" } }
  }'

Delete:

curl -k -u admin:'<password>' \
  -X POST \
  "https://cluster.example.com/api/storage/aggregates" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "aggr_new01",
    "node": { "name": "node01" },
    "block_storage": { "primary": { "disk_count": 10, "raid_type": "raid_dp" } }
  }'

Verify:

curl -k -u admin:'<password>' \
  -X POST \
  "https://cluster.example.com/api/storage/aggregates" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "aggr_new01",
    "node": { "name": "node01" },
    "block_storage": { "primary": { "disk_count": 10, "raid_type": "raid_dp" } }
  }'

CLI Recreate Process

Create the replacement aggregate according to platform standards. In a lab simulator, this might be a small RAID-DP aggregate. In production, follow platform, disk class, RAID size, encryption, and ADP standards.

curl -k -u admin:'<password>' \
  -X POST \
  "https://cluster.example.com/api/storage/aggregates" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "aggr_new01",
    "node": { "name": "node01" },
    "block_storage": { "primary": { "disk_count": 10, "raid_type": "raid_dp" } }
  }'

Validate:

curl -k -u admin:'<password>' \
  -X POST \
  "https://cluster.example.com/api/storage/aggregates" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "aggr_new01",
    "node": { "name": "node01" },
    "block_storage": { "primary": { "disk_count": 10, "raid_type": "raid_dp" } }
  }'

REST API Process

Collect aggregate evidence before destruction:

curl -k -u admin:'<password>' \
  -X POST \
  "https://cluster.example.com/api/storage/aggregates" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "aggr_new01",
    "node": { "name": "node01" },
    "block_storage": { "primary": { "disk_count": 10, "raid_type": "raid_dp" } }
  }'

Delete only after the volume query returns no dependent volumes and the change is approved:

curl -k -u admin:'<password>' \
  -X POST \
  "https://cluster.example.com/api/storage/aggregates" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "aggr_new01",
    "node": { "name": "node01" },
    "block_storage": { "primary": { "disk_count": 10, "raid_type": "raid_dp" } }
  }'

Create replacement aggregate where supported by your ONTAP release:

curl -k -u admin:'<password>' \
  -X POST \
  "https://cluster.example.com/api/storage/aggregates" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "aggr_new01",
    "node": { "name": "node01" },
    "block_storage": { "primary": { "disk_count": 10, "raid_type": "raid_dp" } }
  }'

Best Practices

• Require a second-person review for aggregate deletion.
• Save volume show -aggregate output before deletion.
• Check FabricPool/object-store attachment explicitly.
• Do not reuse names in the same change if it makes audit evidence ambiguous.
• Validate spare disk availability and ownership before recreate.

Backout

After an aggregate is deleted, backout is restore or recreate, not undo. The real rollback is the migration, backup, or replication copy you validated before deletion.