daniel dias SOFTWARE DEVELOPER
← all posts

Azure

Azure Data Box Explained: Why It Exists and Which Option to Choose

Azure Data Box moves large datasets between local environments and Azure when a normal network transfer is too slow, expensive, or disruptive. This article explains how the service works and compares Data Box Disk, Data Box 120, Data Box 525, and Data Box Gateway.

written by Daniel 6 min read
Diagram comparing Azure Data Box Disk, Data Box 120, Data Box 525, and Data Box Gateway for transferring data to Azure

Azure provides several tools for transferring data, including AzCopy, Azure Storage Explorer, Azure Storage Mover, and Data Factory. These tools work over a network connection, which is usually the simplest approach.

The problem appears when the dataset is measured in tens or hundreds of terabytes. Even with a reasonable connection, transferring that volume can take days or weeks, consume most of the available bandwidth, and affect other systems.

Azure Data Box exists for this situation. Instead of sending all the data over the internet, Microsoft ships storage hardware to your location. You copy the data locally, return the device, and Microsoft uploads it to Azure through its datacenter infrastructure. (learn.microsoft.com)

How Azure Data Box works

For an import operation, the general workflow is:

  1. Create a Data Box order in the Azure portal.
  2. Select the destination storage account and provide shipping information.
  3. Receive and connect the device to the local network or computer.
  4. Unlock it using credentials obtained through the Azure portal.
  5. Copy the data to the provided shares or disks.
  6. Return the device to Microsoft.
  7. Microsoft uploads the data to Azure Storage and securely erases the device.

The order and shipment status can be tracked through the Azure portal. Standard Data Box devices also support export operations, where Microsoft copies Azure data to the device before shipping it to you. (learn.microsoft.com)

Data Box is a data transport service. It is not a replacement for Azure Storage, a backup policy, or a continuous synchronization solution. The destination is still an Azure service such as Blob Storage or Azure Files.

Why not transfer everything over the network?

An online transfer may be preferable when the dataset is small, bandwidth is sufficient, or data must move continuously. However, an offline device becomes useful when:

  • The dataset is too large for the available connection.
  • The migration has a fixed deadline.
  • Upload traffic would interfere with production systems.
  • The source location has limited or unreliable connectivity.
  • A large initial copy must be completed before smaller incremental transfers begin.

This last scenario is common in migrations and backup seeding. The initial historical dataset is transferred with Data Box, while later changes use an online transfer mechanism.

Differences between the Data Box options

The product family contains options for different capacities and transfer models.

Option Transfer model Capacity Typical use
Data Box Disk Shipped SSDs Up to 40 TB raw per order, approximately 35 TB usable Smaller offline imports
Data Box 120 Shipped appliance 120 TB usable Large imports or exports
Data Box 525 Shipped appliance 525 TB usable Very large migrations
Data Box Gateway Virtual appliance over the network Depends on local virtual storage and network Continuous or recurring transfers

Data Box Disk

Data Box Disk provides between one and five encrypted 8-TB SSDs. The disks connect through USB or SATA, making them easier to handle than a full appliance. They are intended for importing local data into Azure; export operations are not supported. (learn.microsoft.com)

This option fits below the capacity range of the larger Data Box appliances. It can be appropriate when the network is still the main constraint, but the dataset does not justify ordering a 120-TB device.

Data Box 120 and Data Box 525

These are ruggedized storage appliances designed for larger datasets. Both import and export workflows are supported. Data Box 120 is the practical starting point for transfers above the Data Box Disk range, while Data Box 525 handles migrations involving several hundred terabytes. (learn.microsoft.com)

Microsoft is replacing earlier 80-TB Data Box hardware as next-generation devices become available in more regions. Data Box Heavy has also been retired and is no longer available for new orders. Availability should therefore be checked for the subscription and region before planning a migration. (learn.microsoft.com)

Data Box Gateway

Data Box Gateway is different because Microsoft does not ship a physical device. It is a virtual appliance deployed in a supported virtualized environment. Applications write data through SMB or NFS shares, and the gateway transfers it to Azure over the network. (learn.microsoft.com)

It is suitable for continuous ingestion, cloud archival, and incremental transfers. It does not solve a lack of bandwidth in the same way as an offline Data Box, but it can complement one: use a physical device for the initial bulk transfer and the gateway for ongoing changes.

Choosing the right option

The decision comes down to three questions:

  1. How much data must be moved? Data Box Disk covers smaller multi-terabyte imports, while the 120-TB and 525-TB appliances cover larger migrations.
  2. Is the transfer one-time or continuous? Physical devices fit bulk and point-in-time transfers. Data Box Gateway fits ongoing network transfers.
  3. Is data moving into or out of Azure? Data Box Disk supports imports only, while the larger Data Box appliances support both directions.

Before ordering, also verify regional availability, supported storage targets, file and object limits, required network interfaces, shipping arrangements, and the complete cost of the operation. These details can determine whether Data Box is more practical than an online transfer.

References

Azure Benefits of Azure Application Insights 6 min read Azure Setting Up Azure Monitor Alerts and Action Groups 4 min read Azure RBAC, Policy and Resource Locks: How Azure Governance Fits Together 4 min read