Can a NAS hard drive be read by removing it and plugging it into a computer? The pop-up tells you not to click the initializer, I've tried all three paths

Last month,
helped Xiao Lin move, and he dug out a dual-drive NAS he had saved up during his college days. The machine wouldn't turn on when powered, and inside were two 2TB red drives lying with photos and movies he'd saved up for five or six years. He removed the disk, put it into the hard drive enclosure, and plugged it into his laptop. Windows popped up with a message saying "Initialize the drive before use," and the mouse moved to the confirm key—I pressed down on his hand. With that click, the partition table was rewritten, leaving no room for saving the photos.
computer doesn't just can't read it, it's that it simply doesn't recognize this kind of 'dialect'
NAS drives, whether Synology, Feiniu, or Tekweima, over 90% are formatted to ext4 or Btrfs—these are the file systems of the Linux world. I wrote a dedicated article about the reasons for choosing them back then (How to choose between ext4 and Btrfs). Windows only recognizes NTFS, exFAT, and FAT32 out of the box; ext4 is a foreign language it has never learned before. So when you insert a drive, there won't be any drive letters. In Disk Management, it either shows 'Not initialized' or shows a lump of unrecognizable partitions.
disks are intact, and the data is not lost. It just stands in front of someone who doesn't understand Chinese, and whatever you say, they wave it off.
there's another detail worth knowing: when you plug a NAS drive into a Linux computer and use lsblk, it's often not a partition but a string. The first few small partitions of several hundred MB are the system's private space; the ones that actually store your data are the large partitions that almost fill the entire disk, especially Synology's disk. So don't assume the drive is damaged just because you see a bunch of "unknown partitions"—your photos and movies are perfectly fine lying in the largest partition. The capacity matches the numbers: the large partition shows how big it is, the amount of stuff your drive stored back then gives you a clear idea.
I tested it with a retired drive: a single command line on a Linux computer
to be clear, I dug out a retired 4TB drive (which used to be a storage drive in the NAS, but after replacing the main drive, it was removed and put in an anti-static bag), and plugged it into an old Ubuntu laptop to test it out.
lsblk immediately recognized the property, with clear zoning and layout. Mount -o is a line to attach it, photos, movies, and compressed files downloaded back then are all intact, with all Chinese filenames in their original state. You can even skip mounting—there's a Linux tool called debugfs that can directly import ext4 directories. I tried writing a text file with a Chinese name and then reading it, and the filename, content, and timestamp were exactly the same. The data is just lying on the platter, having nothing to do with whether the NAS machine is dead or not.
note, I always use -o ro for mounting, read-only. The iron rule for data extraction is simple: only look, no writing. If you get a moment of impulse and write back, the original project could have been in chaos right then. If the contents of the tray are especially important, add a backup plan: before you start, use DD to create a mirror image of the whole plate, then do all the work on the mirror image and lock the original tray in the drawer. Spend an extra night to get a foolproof one.
only Windows is at hand: three routes, try them in this order
| plan | whether it costs money | suitable for what situations |
| DiskGenius free version | free | Browsing directories, exporting photo documents, and other small files can be |
| Linux Reader | free | DiskGenius copy with time-limited filling, and you can just drag them out to save them |
| Ubuntu Live USB | is free, | large volume, and holds a single file of several dozen gigabytes—the most reliable route |
Xiaolin's computer is Windows, so I had him install the free version of DiskGenius: once you plug in the drive, you can see the large ext4 partition in the software, and when expanded, you'll see the shared folder from back then. You can preview photos and thumbnails. Over twenty thousand photos were copied in several batches. The middle was stuck on a Blu-ray original drive of over ten GB in the middle, with copying restrictions. Those large files were supplemented using the second solution, Linux Reader. If you really encounter a large number of files or messy files, don't hesitate—just use Ubuntu Live USB: a USB drive as a boot disk, boot from the USB drive into a temporary Linux desktop, plug in the hard drive enclosure, and copy and paste directly. This is the safest route. Linux reads ext4 as your native language, and it won't write anything to your computer's hard drive.
no matter which path you take, if you see any prompt saying "Format the drive before use," click undo it without exception. This disk doesn't need to be formatted; what needs to be read is by someone else.
array disks, splitting them into a single piece is mostly wasted
Xiaolin was lucky. His car was set to Basic mode back then, with each drive independent, so each drive was a complete set of data. But if you had accidentally assembled the array back then, things would have gone awry.
| the model of the group back then | whether a single piece could read |
| Basic or | a single drive, and plugged in would make a complete set |
| RAID1 image | work; any piece is a complete copy. If you have to reassemble it on Linux |
| RAID0 strip | you can't; if the data is split into two parts, missing one won't be complete |
| SHR or RAID5 | is basically out of reach, with only fragments on the single pieces |
RAID1 single drive still can't be read in Windows software, because there's another layer of array metadata pressed on the disk, and Windows doesn't recognize this layer. Back on Linux, use the mdadm tool: first check the array information on the disk, then use mdadm --assemble --run to force it up in a downgraded state due to lack of personnel, and mount -o ro still mounts read-only. Synology disks often have an LVM layer on top of the array, requiring an extra vgchange -ay to activate the logic volume. Commands are everywhere online, but I'll emphasize this once: read only, finish mdadm -- stop and clean it up.
SHR is Synology's proprietary mixed installation solution, structurally more complicated than standard RAID ( I've calculated the difference from RAID1). If you want to read it directly after splitting it into individual units, don't waste time on it—either reinstall the drive back to Synology for full machine read, or accept the loss and recover. When the pallet physically breaks down, that's another story. the opening recovery, I've waded through once, and the price is enough to buy several new units.
want to plug the drive into another NAS and continue using it? There are two scenarios
goes beyond 'reading once' and asks a question further: can the old drive be plugged directly into the new NAS to use as your own?
Synology to Synology, the official route is called HDD migration: upgrade the old machine to the latest DSM version, shut down, insert the disks in the original order into the new machine, and after booting, the new machine automatically reinstalls the system. After installing the storage pool and shared folders, they return as they were. The premise of this path is that both sides are Synology and their models are mutually compatible; cross-brand options are out of reach.
cross-brand direction is "readable, not taken over." An ext4 drive removed from Synology was plugged into Feiniu, which recognized ext4 because it was also a Linux dialect, and an external input could read data and copy it away; But if you want the new NAS to take over this disk along with the original storage pool settings, it can't and shouldn't; each company's array and storage pool carry their own private information. Reading out, moving in, and rebuilding the pool—this is the proper way to migrate across systems. I've written the entire process of relocating and moving , so just follow it.
one more thing you must put here: if the drive is currently in service on the NAS, don't get itchy to pull it out. RAID1 can hold up even if you pull a piece online, at most it will enter degradation mode; RAID0 or single-drive mode offline removal can cause data damage at best, or the entire pool will be scrapped. If you want to disassemble the drive, shut down first—no exceptions.
sit down according to the numbers, and here's the order in which you get started
first understand the condition of the disk before acting, which is more important than any tool:
| your situation | the path to take |
| NAS the machine can still boot | via network copying, SMB or package export, and disassembly is the last option |
| machine won't light up, the disk is Basic or single | Windows starting with DiskGenius, large files converted to Ubuntu Live USB |
| machine won't light up, drive group RAID1 | Ubuntu Live USB with mdadm assembly, read-only |
| drive makes abnormal noises or does not rotate during the entire process | do not power on repeatedly, just observe the recovery method |
Here's a closing tip for
order: After data scraping, if you want to sell second-hand ones, remember to wipe them all before shipping—quick formatting is like tearing up the table of contents pages but leaving the whole book on the shelf, experts flip through it and it's all back. It's just for a temporary computer to read. When hanging a USB hard drive enclosure, pay attention to power supply. For 2.5-inch drives, the box's built-in cable is enough. For 3.5-inch drives, you must have the big brick with independent power supply. I've the pitfall of insufficient power .
Xiaolin's two disks—after copying the photos, one was kept as a cold backup, and I helped him erase and hang it on Xianyu. When it comes to data, machines can break and disks age. Having an extra copy is always more valuable than having an extra trick to save data—when it comes to accidental deletion, recycle bin and snapshots can save a once, but the time it can't be saved depends on the extra copy made today.
