Drive Cages & RAID Enclosures
Multi-bay enclosures get marketed under "RAID" as a category name whether or not they actually do RAID in hardware. That distinction is the single most important thing to check before buying one — and it's usually buried a few lines down the spec sheet, not in the product title.
The distinction that actually matters
A hardware RAID enclosure has a controller that implements the array mapping inside the enclosure rather than relying on the host operating system to combine separately exposed disks. Moving the complete enclosure may preserve that controller-defined layout, but portability still depends on the enclosure, controller, configuration metadata and host compatibility. A "multi-bay DAS enclosure" without that controller, despite often looking identical and sitting in the same product category, just presents each drive to the computer separately; any RAID array has to be built and managed entirely in the operating system's own software instead. Both are legitimate products — a fast multi-drive box for independent volumes is a completely reasonable thing to want — but only one of them is actually a RAID device, and the two examples below, from the same manufacturer, illustrate the difference concretely.
This isn't just a shopping footnote. It connects directly to RAID Array Failure and Failed or Interrupted RAID Rebuild elsewhere in this catalog: recovery depends on knowing which layer created the layout. Hardware controllers, mdadm, Storage Spaces and other software-defined stacks all have their own metadata and failure behavior; none should be assumed inherently more predictable during a damaged-state rebuild.
TerraMaster D4-320
4-bay, no hardware RAIDDespite the "RAID enclosure" category it's sold under, this specific model has no onboard RAID controller — it presents each of its four bays as a separate disk, and any array has to be built in the OS. It is appropriate when separately exposed disks are wanted; any host-managed RAID or storage pool remains dependent on the software stack that created it.
- Bays4, diskless (bring your own drives)
- ConnectionUSB 3.2 Gen 2, 10Gbps; also Thunderbolt 3/4 compatible
- RAIDNone in hardware — single-disk mode only, OS-level RAID only
TerraMaster D5-300
5-bay, hardware RAIDThe genuine article: a dedicated hardware RAID controller with its own configuration software, supporting the standard redundant levels plus a straightforward one-to-one clone mode. The array mapping is managed by the enclosure rather than by host software, but that should not be treated as a guarantee of portability to unrelated hardware.
- Bays5, diskless (bring your own drives)
- ConnectionUSB 3.1 Gen 1, Type-C
- RAIDHardware RAID 0, 1, 5, 10, Clone, JBOD, single disk
Check how the disks are presented to the host and where the array mapping is implemented. Marketing terms such as “RAID,” “RAID-ready,” JBOD and “hardware RAID” are not enough by themselves; verify the supported modes, controller documentation and whether the host sees one logical device or the individual member disks.
For recovery rather than purchasing, see RAID Reconstruction Basics.
For failure analysis, continue with RAID Array Failure, NAS Failure and Understanding Storage Layers Before Recovery.
A useful recovery question is whether the computer sees individual physical members, one controller-defined logical volume, or several LUNs/volumes. Recovery software must work at the layer that actually contains the expected filesystem. Scanning a single RAID member for files can produce fragments while missing the array-level mapping entirely.
What the host actually sees
If an enclosure fails but its disks contain important data, record bay order, model/serial information, configuration screens, logs and controller details before moving members. Do not initialize, create a new array or accept an automatic rebuild merely because replacement hardware offers to “import” or “repair” the set.
The controller may store array metadata on member disks, in enclosure/controller memory, or both. Member order, RAID level, stripe geometry, offsets and controller-specific metadata can all matter. Some bridges or controllers also alter sector presentation or apply transparent encryption. Replacing an enclosure or controller with a superficially similar model is therefore not guaranteed to reproduce the same logical device.
Why the original enclosure can matter in recovery
DAS describes direct-attached storage and says little about how disks are mapped. JBOD is used ambiguously: it may mean individually exposed disks or a concatenated/span mode. A hardware RAID enclosure presents a controller-created logical device. A NAS adds a computer, operating system and network services and may internally use RAID, pools, LVM, ZFS, Btrfs or vendor-specific layers. The label on the chassis is therefore not enough to reconstruct its data layout.
JBOD, DAS, RAID and NAS are different layers
This page describes the general situation, not a guaranteed fix for your specific case. If you'd rather have someone experienced take a look before you try anything, see what a hands-on evaluation looks like.