You can migrate existing volumes to the [Flex Unified](https://docs.cloud.google.com/netapp/volumes/docs/discover/features.md.txt#flex_unified_default-mode_versus_ontap-mode)
service level by using host-based copy tools, backup and restore, or
support-assisted replication. Depending on your volume size and downtime
requirements, you can select the most suitable [migration tool](https://docs.cloud.google.com/netapp/volumes/docs/get-started/migration-to-flex-unified.md.txt#migration_tools)
for your workload.

> [!IMPORTANT]
> **Important:** As the Flex Unified service level is expanding, the Flex File service level is being phased out. Effective September 15, 2026, you can't create new Flex File storage pools, but existing pools remain supported. If you need assistance migrating to Flex Unified, [send an email to the support team](mailto:flex-file-migration@google.com). If you have specific region requirements or need support for Assured Workloads with Flex Unified, [contact the Google Cloud sales team](https://cloud.google.com/contact).

The Flex Unified service level provides all NetApp ONTAP features
on Google Cloud for both file and block storage on a single platform.

The following table compares the features available in the Flex Unified service
level against the legacy service levels (Flex File, Standard, Premium, and Extreme).

| Feature | Flex Unified | Flex File (legacy) | Standard, Premium, and Extreme (legacy) |
|---|---|---|---|
| **Block storage (iSCSI and NVMe/TCP)** | Yes | No | No |
| **Dual-protocol volumes (SMB and NFS)** | Yes | No | Yes |
| **Auto-tiering** | Yes | Yes | Standard: No Premium or Extreme: Yes |
| **ONTAP capabilities (ONTAP-mode pools)** | Yes | No | No |
| **ONTAP integration (SnapMirror and FlexCache)** | Yes | No | Yes |
| **Selective file restore and cross-region backups** | Yes | No | Yes |
| **Shared QoS at the pool level** | Yes | Yes | No |
| **Thin clones** | Yes | No | No |
| **Large capacity pools and volumes** | Yes | No | Standard: No Premium or Extreme: Yes |
| **Throughput** | Increased throughput with large volumes | Standard throughput | Increased throughput with large volumes |
| **Regional availability** | Yes | Yes | No |

## Migration tools

You can use the following tools to migrate volumes to the Flex Unified service
level:

- [Host-based data copy](https://docs.cloud.google.com/netapp/volumes/docs/get-started/migration-to-flex-unified.md.txt#host-based_data_copy)

- [Backup and restore](https://docs.cloud.google.com/netapp/volumes/docs/get-started/migration-to-flex-unified.md.txt#copy_data_using_backup_and_restore)

- [Support-assisted replication](https://docs.cloud.google.com/netapp/volumes/docs/get-started/migration-to-flex-unified.md.txt#copy_data_using_support-assisted_replication)

Use the following table to select the appropriate tool for your use case.

| Tool | Use case | Considerations |
|---|---|---|
| Host-based data copy | Reorganizing volume data, either by removing or consolidating directories, or moving data from a regular volume to a large capacity volume | - Mount IP address changes |
| Backup and restore | Moving a few small volumes | - Restore at 300 GiB per hour - Mount IP address changes |
| Support-assisted replication | Moving a few large volumes | - Preserves snapshots and SMB permissions - Supports immediate cutover after final incremental update - Mount IP address changes |

## Host-based data copy

To move data to Flex Unified volumes by using host-based copy tools, complete
the following steps:

1. Select one of the [data copy tools](https://docs.cloud.google.com/netapp/volumes/docs/get-started/migration-to-flex-unified.md.txt#data_copy_tools) described in this
   section to create a full copy of the data.

2. [Create a full copy of the data](https://docs.cloud.google.com/netapp/volumes/docs/get-started/migration-to-flex-unified.md.txt#copy_data_using_host-based_tools), and then
   run incremental copies to capture any changes.

### Data copy tools

Select a data copy tool based on your client operating system and protocol:

- [Robocopy](https://docs.cloud.google.com/netapp/volumes/docs/get-started/migration-to-flex-unified.md.txt#robocopy): Copies SMB files on a Windows VM.

- [NetApp XCP](https://docs.cloud.google.com/netapp/volumes/docs/get-started/migration-to-flex-unified.md.txt#netapp_xcp) or rsync: Copies NFS files on a Linux VM.

#### Robocopy

Robocopy is a [Microsoft file copy tool](https://learn.microsoft.com/en-us/windows-server/administration/windows-commands/robocopy)
for Windows systems.

Consider the following options when you use robocopy:

- `/mir`: Copies all subdirectories, including the empty subdirectories.

- `/sec /secfix`: Copies security information, such as access control lists
  (ACLs), including information about the unchanged files with security changes.
  For more information, see [robocopy mirroring file permissions](https://techcommunity.microsoft.com/blog/filecab/robocopy-mir-switch-8211-mirroring-file-permissions/423662).

- `/z`: Enables restartable mode, which allows robocopy to resume if
  interrupted.

- `/mt:32`: Uses 32 threads for multi-threaded copies.

- `/b`: Uses backup mode to override file and directory ACLs. Robocopy users
  must be members of the backup operators group on both the source and
  destination volumes.

- `/r:2 /w:1`: Retries failed copies with a one-second wait.

- `/log:<var>logfile</var>`: Logs robocopy status output.

#### NetApp XCP

XCP is a [NetApp data migration tool](https://xcp.netapp.com/) that copies files
during data migrations. You can use it to migrate data to Flex Unified service
level volumes. For additional assistance in using this tool, contact your NetApp
sales specialist.

### Copy data using host-based tools

To copy data using a host-based copy tool, complete the following steps:

1. Create a NetApp Volumes Flex Unified [pool](https://docs.cloud.google.com/netapp/volumes/docs/configure-and-use/storage-pools/create-storage-pool.md.txt)
   and a [volume](https://docs.cloud.google.com/netapp/volumes/docs/configure-and-use/volumes/create-volume.md.txt).

2. [Mount](https://docs.cloud.google.com/netapp/volumes/docs/get-started/quickstarts/mount-volume.md.txt#mount_your_volume_on_a_client)
   the existing volume and the new Flex Unified volume to a virtual machine (VM)
   in your Virtual Private Cloud (VPC) network. Use a
   [data copy tool](https://docs.cloud.google.com/netapp/volumes/docs/get-started/migration-to-flex-unified.md.txt#data_copy_tools) to transfer the data to the Flex Unified
   volume.

3. To perform a final incremental copy of the data to the Flex Unified volume,
   disconnect clients from the current volume during scheduled maintenance.
   Verify that you have copied all files by checking file counts, checksums, and
   similar validation methods.

4. Connect the clients to the Flex Unified volume and test access.

## Copy data using backup and restore

The [backup and restore](https://docs.cloud.google.com/netapp/volumes/docs/protect-data/about-backups.md.txt)
feature in NetApp Volumes is available with Standard, Premium,
Extreme, Flex File, and Flex Unified service levels. You can use this feature to
back up a volume and restore it to a new Flex Unified storage pool.

### Considerations

Consider the following details when you migrate using the backup and restore
feature:

- NFS exports and SMB shares receive a new IP address. For SMB shares, the
  NetBIOS name retains its prefix and receives a new five-character suffix.

- You can't use hyphens in Flex Unified volume names, but you can still use
  hyphens in export and share names. These volume name changes can affect
  Terraform, other automation, and your monitoring and alerting configurations.

- The restore process migrates SMB permissions and NFS ACLs along with your
  files and folders.

- Backup and restore doesn't migrate existing snapshots. For the new volume, you
  must create export and tiering policies, and data protection policies for
  snapshots, backups, and replications.

- Backup features, such as cross-region backups and customer-managed
  encryption keys (CMEK), aren't supported with Flex File, and you can't use
  them during migrations.

- The Flex File service level doesn't support large capacity volumes. You can't
  use the backup and restore feature to migrate data from a Flex File volume to
  a large capacity Flex Unified volume. Use a host-based data copy instead.

- You can't choose a specific time for scheduled backups. If you have scheduled
  backups configured, check when previous backups ran to estimate the next
  backup time. For more control over backup timing, use manual backups.

- Billing for the original pool continues until you complete the migration and
  delete the pool.

### Migrate a volume using backup and restore

To migrate a volume using the backup and restore feature, complete the following
steps:

1. Create a baseline backup of the Standard, Premium, Extreme, or Flex File
   volume before migration. If the volume has scheduled backups, you can use one
   of them. This backup serves as a baseline and helps reduce the time needed
   for the final incremental backup.

2. [Create a NetApp Volumes Flex Unified pool](https://docs.cloud.google.com/netapp/volumes/docs/configure-and-use/storage-pools/create-storage-pool.md.txt)
   if needed.

3. During a maintenance window, complete the following steps:

   1. Disconnect the clients from the existing volume and perform a final
      incremental manual backup of the data.

      After the backup status changes to **Ready**, wait 20 minutes for the
      process to finish, and then continue to the next step.
   2. After the incremental backup completes, select
      **Create volume from backup** using the latest incremental backup, and
      then choose the Flex Unified storage pool where you want to create the new
      volume.

      > [!NOTE]
      > **Note:** For planning purposes, assume a restore rate of approximately 300 GiB per hour. You can track the progress in **Notifications** in the Google Cloud console.

   3. After the restore completes, connect a client to the Flex Unified volume,
      and then verify the volume using your application.

4. Connect your remaining clients to the Flex Unified volume.

## Copy data using support-assisted replication

The support team can help you replicate Flex File volumes to new Flex Unified
volumes. First, the team creates a baseline copy of your volumes, and then
copies subsequent changes through smaller incremental updates.

When you're ready to move clients to the new volumes, you must schedule a
maintenance window. During this window, you disconnect the clients, run a final
incremental update, and then reconnect the clients to the new volumes. This
approach minimizes downtime because the final update transfers only recent
changes.

### Considerations

Consider the following details when you migrate using support-assisted
replication:

- NFS exports and SMB shares receive a new IP address. For SMB shares, the
  NetBIOS name retains its prefix and receives a new five-character suffix.

- Replication replaces hyphens in volume names with underscores, but you can
  still use hyphens in export and share names. These volume name changes can
  affect Terraform, other automation, and your monitoring and alerting
  configurations.

- Replication copies SMB permissions and NFS ACLs along with your files and
  folders.

- Replication copies existing export, snapshot, and tiering policies to the new
  volumes. To protect data on the new volumes, you must create backup and
  replication policies.

- Replication copies existing snapshots to the new volumes.

- Billing for the original pool continues until you complete the migration and
  delete the pool.

### Migrate volumes using support-assisted replication

To migrate volumes using support-assisted SnapMirror replication, complete the
following steps:

1. In your project, identify the Flex File pool and volumes that you want to
   migrate. The support team reviews these resources because any existing
   backups or replications must be paused during the migration. If the existing
   volumes use auto-tiering, you can also enable it on the Flex Unified pool.

2. If you have multiple storage volumes, decide how to group them for migration
   to minimize client mount changes. Consider the following factors when you
   group volumes:

   - The original IP address for the volumes changes to a new address from the
     `/28` CIDR range used for the Flex Unified pool.

   - Flex Unified volumes require underscores instead of hyphens in
     Google Cloud volume resource names. However, the share and mount paths
     remain the same. For more information, see
     [Volume and share naming rules](https://docs.cloud.google.com/netapp/volumes/docs/configure-and-use/volumes/create-volume.md.txt#volume_and_share_naming_rules).

   - Each pool supports up to 10 concurrent SnapMirror replications.

3. [Create a NetApp Volumes Flex Unified pool](https://docs.cloud.google.com/netapp/volumes/docs/configure-and-use/storage-pools/create-storage-pool.md.txt)
   based on guidance from the support team, including any required settings for
   custom performance and auto-tiering. If you need SMB, create a small
   placeholder SMB volume in the pool. If you use auto-tiering, enable hot-tier
   bypass during the migration.

4. Work with the support team to create the new Flex Unified volumes and
   SnapMirror replications to start the baseline data copy.

   The baseline copy typically runs at about 1 TiB per hour and can reach up to
   2.6 TiB per hour, or 764 MiBps, on pools with higher custom performance
   settings. Because replication prioritizes client I/O, the replication speed
   might decrease depending on workload activity. After the baseline copy
   completes and replication switches to incremental updates, the volumes are
   ready for cutover.
5. During a maintenance window, complete the following steps:

   1. Disconnect the clients from the original volumes.

   2. Work with the support team to run a final incremental update, and then
      stop replication.

   3. Confirm that you can view and manage the Flex Unified volumes in your
      project.

   4. Verify that the new volumes inherited the export, snapshot, and tiering
      policies, along with the NFS ACLs and SMB NTFS permissions, from the
      original volumes. Replication automatically copies these settings with
      your files.

   5. Connect a client to the Flex Unified volumes, and then verify access using
      your application.

6. Connect your remaining clients to the Flex Unified volumes.

7. Recreate any required backup and replication policies on the new volumes.

## What's next

For more information, see the following resources:

- [Connect SMB clients](https://docs.cloud.google.com/netapp/volumes/docs/connect-clients/connect-smb-clients.md.txt).

- [Connect NFS clients](https://docs.cloud.google.com/netapp/volumes/docs/connect-clients/connect-nfs-clients.md.txt).

- [Connect iSCSI clients](https://docs.cloud.google.com/netapp/volumes/docs/connect-clients/connect-iscsi-clients.md.txt).

- [Connect NVMe/TCP clients](https://docs.cloud.google.com/netapp/volumes/docs/connect-clients/connect-nvme-tcp-clients.md.txt).