NAS Rebuild Failed After Disk Replacement? Stop and Call First.
UK NAS and RAID recovery specialists since 2002. All brands: Synology, QNAP, Buffalo, WD, Drobo, NetGear. SHR, RAID 5/6, Btrfs, ext4. Free UK collection. No-fix-no-fee.
What this means and what to do next
A NAS rebuild that starts and fails is one of the most common data-loss scenarios we handle. The sequence of events is almost always the same: a disk is reported as failed, the user correctly replaces it with a new disk and starts the rebuild, and part-way through the rebuild a second disk reports errors and the rebuild aborts. The volume goes offline. The NAS shows an error message — "Volume crashed", "Rebuild failed", "Degraded array", or a blinking status light depending on the brand.
What's happened is not a coincidence: the rebuild process reads every sector on every surviving disk for the first time in potentially years. If any of those disks has accumulated unreadable sectors over time — sectors that RAID has been silently compensating for by calculating the missing data from parity, but has never actually had to read — those sectors are exposed during the rebuild read-sweep. The NAS controller cannot complete the parity reconstruction and drops the second disk, taking the array offline. The disks are typically not dead. They have specific regions of bad sectors, and everything outside those regions is intact.
The critical mistake at this point is to retry the rebuild, or to insert yet another replacement disk and let the NAS try again. Each rebuild attempt performs a massive sequential write across the array based on whatever parity it has reconstructed so far — and if that reconstruction is incomplete (because of the bad sectors), the written parity is wrong. The next thing that reads that data will get garbage. We see drives every week where a clean imaging job has been complicated by multiple rebuild attempts that overwrote good data with incorrect parity.
The three most common causes of a NAS rebuild failing
1. Latent bad sectors on a surviving disk — revealed by the rebuild read. By far the most common scenario. RAID 5, RAID 6, and Synology SHR all use parity to reconstruct data from a failed disk. When they reconstruct, they read every sector on every other disk in the array. A disk that has been accumulating uncorrectable read errors for years — but whose errors have been masked by the redundancy — will fail these reads during the rebuild sweep, and the array controller drops that disk too. The solution is to image all disks first using specialist hardware (PC-3000 or Atola DiskSense) with extended retry and error-correction that the NAS controller's native rebuild cannot do, then reconstruct the array from the images.
2. Replacement disk is too small, incompatible, or failed immediately. A replacement disk that is slightly smaller than the original (a common issue with drives from different manufacturers at the same nominal capacity), uses a different sector size (4K versus 512e), or fails early in its life will abort the rebuild. If the replacement disk itself has early-life failures — infant mortality is a real phenomenon in consumer drives — it may have overwritten the first few gigabytes of the array stripe before failing. Recovery must account for the overwritten sectors.
3. NAS firmware corruption during or after the rebuild attempt. Some NAS firmware updates install during a rebuild or are triggered by the same reboot that started the rebuild. A power cut mid-rebuild can corrupt the NAS filesystem metadata (Btrfs or ext4) even before the RAID layer is damaged. In these cases the disks themselves are fine and the RAID is intact, but the filesystem can't be mounted. Recovery at the filesystem layer (rather than the RAID layer) is faster and less expensive for this specific scenario.
How Data Clinic recovers a NAS after a failed rebuild
The first step is always to image all disks individually — including both the originally-failed disk and any replacement disks that were inserted, even if they appear healthy. Specialist imaging hardware (PC-3000 or Atola DiskSense) applies extended retry passes and hardware-level error correction to read bad sectors that the NAS controller gives up on; in most cases we recover 99% or more of sectors across the array, including the regions that caused the rebuild to fail. All original disks are returned untouched after imaging.
From the images, we reconstruct the RAID geometry virtually. For Synology SHR and QNAP RAID, the on-disk metadata tells us the disk order, stripe size, and parity rotation. For Buffalo and other brands with proprietary RAID implementations, we determine the layout from the data patterns. We then assemble the logical volume and apply filesystem recovery — Btrfs for modern Synology, ext4 for older units and most QNAP, HFS+ or APFS for some NAS units with Apple-specific firmware.
Once the filesystem is readable, we extract and verify your files. For business customers with live systems, we can structure the output to match your existing share layout. We handle mixed-media arrays (arrays where some disks were replaced partway through their life and have different bad-sector maps), arrays that were partially rebuilt before the failure was caught, and arrays where the NAS itself is no longer functional but the disks are intact. More about our RAID and NAS recovery service →.
Get a free initial diagnosis in 60 seconds
In the tool below, choose RAID / NAS → Rebuild failed or volume crashed for the right recovery path.
What our customers say
"Three years of family photos on a drive that suddenly failed. Data Clinic collected next day, kept me updated through the cleanroom work, and got everything back. Worth every penny."
"Honest, fixed-price, no-fix-no-fee. Quoted by another lab at three times the price. Recovered 100% of my files."
"Reasonable cost, clear communication, and they were straight with me about what was recoverable and what wasn't. Recommended."
Frequently asked questions
Should I label the disks with their bay numbers before removing them?
Yes — this is the single most important thing you can do before sending a NAS to any recovery lab. The disk order in the array directly affects which sectors hold which stripes, and getting it wrong means every file is corrupted. Mark each disk with a permanent marker (Disk 1 / Bay 1, Disk 2 / Bay 2, etc.) before removing. If you've already mixed them up, send the NAS too — we can usually determine disk order from the on-disk metadata.
How much does NAS rebuild failure recovery cost in the UK?
Most cases fall in our standard RAID tier: typically £695 to £1,495 including VAT depending on the number of disks, total capacity, the NAS brand and filesystem (Synology Btrfs and QNAP are the most common; proprietary formats like Buffalo's may be higher), and the extent of damage from the rebuild attempt. Emergency turnarounds for business customers are available. Free diagnosis. No fee if we cannot recover your data.
The NAS says the originally-failed disk is still failed and the new replacement is also failed. Both are still in the NAS. What should I do?
Power the NAS off and leave it off. Don't remove or swap any more disks. Label the bay positions on the NAS chassis if you can, then contact us. We need both the originally-failed disk and the replacement, plus all surviving disks. The "failed" disks are often the most important source of data for the reconstruction.
The rebuild was already 70% complete when it failed. How much is lost?
Much less than you'd fear. The 70% that was written during the rebuild may itself be correct if the second disk didn't fail until late in the process. From the images, we can determine at exactly what sector the rebuild aborted, take the parity data that was correctly written up to that point, and combine it with the data from the original disks for the remaining 30%. In many partial-rebuild cases we recover 95 to 100% of the files.
All brands — Synology, QNAP, Buffalo, WD MyCloud, Drobo, NetGear ReadyNAS?
Yes. We handle all of them. Synology SHR and QNAP RAID are our most common NAS cases. Buffalo LinkStation and TeraStation use a variant of RAID 5 that we've worked with extensively. WD MyCloud Mirror and Pro use a JBOD or RAID 1 layout that's generally simpler. Drobo uses its own proprietary BeyondRAID filesystem, which is not standard RAID — recoverable, but requires specific forensic tools. We handle all of these routinely.
How long does NAS rebuild failure recovery take?
Standard turnaround is 7 to 14 working days from receipt of the disks. Emergency 48-hour and 72-hour services are available for business customers — call 0800 151 2207 to discuss.