Synology NAS Data Recovery: When Drives Fail at Home or in the Office
Our founder and lead technician, Joseph Montanti, handles cases personally.
If your Synology NAS has a disk that clicks, beeps, grinds, disappears or returns read errors, shut it down. If the remaining disks sound normal and read steadily, File Station may still open a crashed volume, and a degraded two-drive RAID 1 pool may still serve shared folders from its surviving mirror. Synology gives different instructions for readable and unreadable crashed volumes.
On this page
Crashed volume or degraded storage pool?
Degraded means a redundant pool has lost some of its protection but can still operate with the remaining drives. Crashed means the volume or pool can no longer operate normally. Drive errors, missing members and file-system errors can all cause a crash. On a DS920+ or DS1522+, the exact label in Storage Manager matters more than the color of the front light.
DSM is the management software. A storage pool combines physical disks, a volume holds a file system on that pool, and shared folders live in the volume. You can sign in to DSM even when the volume is inaccessible.
Stop any repair or rebuild already running on a crashed pool and power the NAS off. Shut it down if the volume is unreadable or a copy stalls. If two drives have failed in SHR-1 or RAID 5, shut it down and do not insert a replacement because the pool has lost more than it can rebuild.
If the NAS must stay off for RAID data recovery, label each disk with its bay number before removal and keep every original, including the failed disk. For mail-in recovery, pack each removed disk in an antistatic bag and cushion it in a rigid box.
- For a quiet, steadily readable volume with no repair underway, connect a computer on the same local network. Choose a destination on that computer or on attached storage with enough free space, outside the affected pool.
- Use the working share or File Station to copy your current work, then less urgent folders. Keep the pool's existing drives in place during the copy. Open a few of the files from the destination and check the latest versions.
Leave Repair, Remove, Create, volume repair and data scrubbing alone while these disks hold your only copy. A snapshot on that pool shares the same disks.
RAID 1, SHR and parity after a disk drops out
| Layout before the fault | What losing a member changes |
|---|---|
| Basic, JBOD or RAID 0 | There is no redundant copy from which DSM can reconstruct a failed disk. A striped file may need pieces from more than one member. |
| Two-drive RAID 1 | Each member holds a mirror of the volume. A surviving current member can keep the share available after one fails. |
| SHR with one-drive protection | One missing member is within the configured protection. A second loss before repairs finish can take the pool offline. |
| SHR-2 or RAID 6 | The layout provides for two simultaneous drive failures. Further missing data can exceed what parity can reconstruct. |
| RAID 5 | One disk's missing blocks can be calculated from the others. Unreadable blocks on a surviving disk complicate that calculation. |
| RAID 10 | The identity of the failed disks matters. Losing both copies of one mirrored pair can make part of the striped volume unreadable. |
Synology Hybrid RAID, or SHR, can combine regions of different-sized disks into a single pool. A one-drive SHR pool has no redundancy. Two drives can provide a mirror. Larger SHR arrangements can use parity, depending on the configuration. Give the configuration shown in Storage Manager more weight than the number of trays. Synology's standard DSM list includes RAID 5 with distributed parity, alongside SHR, RAID 6 and other choices. Synology's RAID type table gives the minimum counts and failure tolerance.
A mirror is a second copy of the state of the volume. If a folder was deleted yesterday, a healthy mirror normally reflects yesterday's deletion as well. If one disk fell out of the mirror before someone changed the folder, the two members can instead represent different points in time. That older disk is worth keeping. Rebuilding onto it trades its previous contents for a new copy of the surviving member.
What DSM repairs when you replace a disk
Check the failed bay against Storage Manager and the DSM log before pulling a drive. For a degraded, redundant pool with a verified backup, DSM's repair operation writes the missing member's contents to a replacement. It has to read the surviving disks to do that. Let a healthy rebuild finish before shutting down. It cannot reconstruct an old version of a file that was already missing, and it can't repair a bad directory merely by making another copy of it. Synology's pool repair directions specify compatible drive types and capacities for the pool you have.
If a drive came out of the wrong bay, keep it with the other originals and label the replacement separately. It may hold only part of a rebuild. Leave the disks as they are before making another swap.
A striped layout explains why a second read failure matters. Imagine a stripe with one missing data block. The other data blocks and that stripe's parity may be enough to calculate it. If another required block in the same stripe can't be read, a single parity value can't fill both gaps. Elsewhere on the disks all the pieces of a different file may be readable. The result depends on where the gaps fall as well as how many disks have warning lights.
A DiskStation that won't boot after a power loss or update
If the unit never reaches DSM, the enclosure, system partition and data volume are separate possibilities. A blinking power light or “Not installed” in Synology Assistant is no reason to initialize disks that held a live volume yesterday. New installation creates storage structures. The immediate need is access to the existing structures and the files they describe.
Synology has a supported drive migration procedure, but compatibility depends on the source and destination models and configuration. A newer empty DiskStation isn't automatically the right destination. Check the model-pair requirements, any read-write SSD cache and the recovery key for an encrypted volume before treating migration as an option. If the replacement offers a new installation, initialization, erase or format instead of migration, stop and leave the disks untouched. Reserve migration and other changes for after there is an independent copy. Synology's HDD migration instructions list those conditions.
For a failed enclosure with stable disks, Synology’s Linux PC procedure assembles supported sets and mounts the volume read-only. Never mount the only original members read-write or accept a Windows initialization prompt. A USB dock with one disk from a four-drive SHR pool lacks the other pieces.
The PC procedure has exclusions, including configurations using a read-write SSD cache. Keep any cache devices with the NAS instead of treating them as disposable accelerators. If the pool extends into an expansion chassis, its members belong to the same set. Keep disks from a separate backup pool apart. The DX517 specifications identify models, including DS920+ and DS1522+, that can span a volume across units.
On DSM 7, Online Assemble is a distinct operation for a pool whose existing members DSM can recognize. One situation Synology describes is a previously working pool marked Missing after a live drive was removed by mistake. A compatible replacement disk has no copy of that member's data. The disk that came out of the bay may be the one DSM needs to see again. Preserve that original before treating the warning like an ordinary one-disk failure. Synology's Online Assemble conditions.
If DSM sees too few original members, or one of them develops read errors, the menu choice cannot provide the missing data. Work from the existing backup or the original set as it stands.
Why an assembled array may still have unreadable folders
Reconstructing the disk layout supplies a logical volume. Btrfs or ext4 then has to find the directories and the blocks that belong to each file. Damage to that metadata can leave DSM showing a volume while File Station gives an input/output error for one shared folder. With Btrfs, a snapshot provides an earlier view of the same underlying storage. It can help with deletion if accessible, but damage to shared blocks can affect several snapshots.
An encrypted shared folder adds another requirement. Having the disks and seeing the folder name isn't equivalent to having the key to its contents. Keep the saved folder key or password and any separate volume recovery key available. If a crashed volume is still reachable, Synology publishes a specific route for mounting an encrypted share with its key. Synology describes the prerequisites.
For a virtual machine or iSCSI workload, the needed object may be a virtual disk rather than a folder of ordinary documents. Recovering the outer file or logical unit is one step. The guest file system inside must also be consistent enough to read. If you only need a few files from the guest, see whether a working backup of the VM already contains them before paying to reconstruct all the NAS storage.
If the shared folder is readable and just one file is missing, use the snapshot for that folder before considering a pool repair. In Snapshot Replication, select the shared folder and browse a snapshot from before the deletion. File Station opens that dated view. Copy the file you need to a separate destination and compare its date there. Through an SMB share on Windows, Previous Versions can expose the same earlier views when the administrator has enabled them. Rolling back the entire shared folder to retrieve one spreadsheet would also roll back everyone else's later work. Synology's individual-file snapshot restore.
When to weigh a recovery quote
| Synology configuration | Recovery price | Work that changes the estimate |
|---|---|---|
| Single-drive or one-bay NAS | $1,000 to $3,000+ | Reading one disk, including mechanical work if needed |
| Two-bay RAID 1 or SHR | Quoted per case | Which mirror has the needed version and which areas still read |
| SHR or RAID 5 with missing members | Quoted per case | Reading the surviving pieces and reconstructing the correct layout |
| Larger SHR-2, RAID 6 or expanded array | Quoted per case | Number and condition of all required members |
| Failed enclosure or interrupted rebuild | Quoted per case | Identifying a consistent set and whether a file system can be opened from it |
Open a Hyper Backup archive without rebuilding the NAS
If you backed up shared folders to USB, another NAS or a supported cloud destination, check the most recent completed backup. A multiversion Hyper Backup destination uses an .hbk repository. Use Hyper Backup Explorer or Hyper Backup to browse its versions. An encrypted task needs its backup password or encryption key. Select a date before the loss and open the particular folder you need.
Restore to storage separate from the failed volume. A current file can be overwritten if an older version is restored over it, so check the destination as well as the backup date. You may find that the archive includes all the shared documents but misses one application or last night's changes. That tells you exactly which part of the original array is still worth recovering. Synology's Hyper Backup restore instructions describe the archive formats and restore options.
Get help with data recovery
Bring your device to our Lincroft lab during opening hours, or mail it in.
How to mail your deviceThe $100 diagnostic is non-refundable once diagnostics start, and donor drives or replacement parts you approve are non-refundable once ordered, even if recovery fails.
Published by MDrepairs
MDrepairs is a data recovery and electronics repair lab in Lincroft, New Jersey. Our founder and lead technician Joseph Montanti films real cases for our YouTube channel, showing tools, repair steps and recovery outcomes.
Keep reading
- Windows Storage Spaces Failure: Preserve the Pool Before Resetting Disks
- Synology Volume Crashed: What to Do While Files Are Still Readable
- RAID 1 Recovery: Why a Mirror Does Not Save Every Deleted File
