Migrate data into Azure File Sync with Azure Data Box
Here’s a concise, structured summary of the article, keeping original meaning and links intact.
Overview
Purpose: Use Azure Data Box to bulk-migrate data from an on-premises Windows Server (2012 R2 or newer) into Azure file shares, then optionally enable Azure File Sync (AFS) on the original server so the server becomes a cached (tiered) copy of the cloud.
Scenario applicability:
This guide is for cases where you will install AFS on the same Windows Server that holds the original files and use Data Box for the initial bulk upload.
Not for sources where you cannot install AFS (e.g., NAS) — see Migrate from NAS via Data Box to Azure File Sync or the migration overview page.
Migration route: Windows Server ⇒ Data Box ⇒ Azure file share ⇒ Azure File Sync back to the original server location.
Migration phases (summary)
Phase 1 — Plan mapping and number of Azure file shares
You must keep the original server’s direct-attached storage (DAS) as the source.
AFS sync groups: each sync group ties one cloud endpoint (Azure file share) and at least one server endpoint (server path).
Limits & best practices:
One server can sync up to 30 Azure file shares.
Aim to keep items per share well below scale limits (recommended < 20–30M items; tested to 100M).
Consider share grouping (combine many local SMB shares into a single cloud share as subfolders) or volume-root sync when appropriate.
Map folders to cloud shares in a deployment mapping table (template available: Azure File Sync - Namespace Mapping.xlsx).
Important: finalize folder structure before migration — restructuring during migration undermines the Data Box benefits.
Phase 2 — Deploy Azure storage resources
Provision storage accounts and Azure file shares according to your mapping from Phase 1.
Consider performance (IOPS / throughput) and storage-account limits — best practice: place very active shares in their own storage account.
Make sure storage account regions match the Storage Sync Service region.
Note file-share size limits and redundancy options; follow Create an Azure file share guidance when creating large (100 TiB) shares.
Phase 3 — Choose Data Box options & count devices
Any Data Box appliance can write to up to 10 storage accounts.
Common options:
Data Box Disk: 1–5 SSDs (8 TiB each, up to ~40 TiB physical).
Data Box (appliance): ~80 TiB usable (SMB-based copy only).
Plan which Data Box device will upload which storage accounts and avoid splitting a single file share across multiple Data Boxes.
Phase 4 — Copy files onto the Data Box (seeding)
Set up and connect the Data Box (links: Set up Data Box, Set up Data Box Disk).
Use Robocopy (recommended) rather than Data Box copy tools to preserve file fidelity.
Data Box provides pre-provisioned SMB shares. For standard storage accounts, use shares ending with _AzFiles; ignore blob shares.
Follow Data Box upload preparation steps (connect → copy → prepare for upload).
Recommended Robocopy sample command:
robocopy <SourcePath> <Dest.Path> /MT:20 /R:2 /W:1 /B /MIR /IT /COPY:DATSO /DCOPY:DAT /NP /NFL /NDL /XD "System Volume Information" /UNILOG:<FilePathAndName>Key Robocopy switch meanings explained (e.g., /MIR, /B, /COPY, /DCOPY, /MT, /LFSM, /L for test runs).
Recommendation: use Windows Server 2022 or ensure Windows Server 2019 has latest relevant patches (e.g., KB5005103).
Phase 5 — Deploy Storage Sync Service (cloud resource)
Wait until Data Box data is ingested to Azure.
Deploy a Storage Sync Service in the same region as your storage accounts; one Storage Sync Service is recommended per set of servers/shares that may interact.
Use the deployment guidance in the Azure File Sync deployment article (deploy the Storage Sync Service section).
Phase 6 — Install Azure File Sync agent on the server
Install PowerShell modules first:
Install-Module -Name Az -AllowClobber
Install-Module -Name Az.StorageSyncTurn off Internet Explorer Enhanced Security Configuration before installation if instructed.
If a proxy or restricted firewall is used, configure it or use the agent’s proxy settings; a network connectivity report lists required endpoints after registration.
Download latest agent from: https://aka.ms/AFS/agent
After installation and registration, verify the server appears under Registered servers in the Storage Sync Service resource.
Phase 7 — Configure AFS on the existing Windows Server and enable authoritative upload
Steps to add a sync group and server endpoint:
Create one sync group per Azure file share (cloud endpoint).
In the sync group, Add Server Endpoint for the local folder that matches the cloud folder structure.
Use the checkbox in the Create server endpoint wizard to enable Authoritative upload only if the server folder path exactly matches the folder hierarchy already uploaded to the Azure file share. This sets Initial sync to “Authoritatively overwrite files and folders in the Azure file share with content in this server’s path.”
If folder hierarchies mismatch, enabling authoritative upload will delete cloud content and reupload from server — avoid mismatches.
Optionally enable cloud tiering per server endpoint to keep full namespace available locally while tiering cooler data to the cloud.
Complete the migration
After creating server endpoints, sync will enumerate cloud content and converge changes — enumeration can take long for large namespaces.
Users continue working on the on-prem server throughout, enabling a zero-downtime migration.
Repeat sync group / server endpoint creation for all shares mapped in Phase 1; migration is complete when all server endpoints are created and syncing.
Key cautions & tips
Finalize file/folder structure before starting — restructuring during migration can cause data loss or negate Data Box benefits.
Ensure folder hierarchy on server matches what was uploaded to the Azure file share if you plan to use the Authoritative upload option.
Keep items-per-share and storage-account IOPS/throughput considerations in mind when mapping shares to storage accounts.
Robocopy flags matter for fidelity, performance, and safety; verify command options and use test runs.
One server can sync up to 30 Azure file shares; consider grouping, additional servers, or additional storage accounts to scale.
Useful links from the article (unchanged)
Migration overview: https://docs.azure.cn/en-us/storage/files/storage-files-migration-overview
NAS migration via Data Box: https://docs.azure.cn/en-us/storage/files/storage-files-migration-nas-hybrid-databox
Data Box setup: https://docs.azure.cn/en-us/databox/data-box-quickstart-portal
Data Box Disk setup: https://docs.azure.cn/en-us/databox/data-box-disk-quickstart-portal
Data Box documentation overview: https://docs.azure.cn/en-us/databox/data-box-overview
Create an Azure file share: https://docs.azure.cn/en-us/storage/files/storage-how-to-create-file-share
Azure File Sync deployment guide: https://docs.azure.cn/en-us/storage/file-sync/file-sync-deployment-guide
Azure File Sync agent download: https://aka.ms/AFS/agent
Cloud tiering overview: https://docs.azure.cn/en-us/storage/file-sync/file-sync-cloud-tiering-overview
Cloud tiering policies: https://docs.azure.cn/en-us/storage/file-sync/file-sync-cloud-tiering-policy
Azure Storage redundancy: https://docs.azure.cn/en-us/storage/common/storage-redundancy
Data Box Disk docs: https://docs.azure.cn/en-us/databox/data-box-disk-overview
If you’d like, I can:
Produce a one-page checklist you can use during the migration (phases with actionable items).
Convert the Robocopy recommendation into a templated command you can fill with your paths.
Was this helpful?