First: If a RAID or NAS contains the only copy of important files and its state is uncertain, pause changes and involve the person responsible for the system. Do not initialise an array or start a speculative rebuild.
A failed shared folder can stop a household, a small office or a larger business. The right response depends on the system state and available copies—not just the amount of storage.
Degraded is not the same as crashed
A degraded array may still serve data while lacking part of its normal redundancy. A crashed or inaccessible volume is a different situation. Read the exact manufacturer status and record it. Do not treat generic advice about replacing one failed disk as instructions for every array.
Synology documents several causes of crashes, including exceeding the RAID fault tolerance and file-system errors. That is why the number of disks alone does not identify a safe recovery procedure.
- 01
Record
Status, errors and drive positions.
- 02
Preserve
Agree how to prevent further changes.
- 03
Assess
Check the copies and specialist options.
- 04
Validate
Have the data owner check the result.
Record the system before changing it
- NAS/server make, model, operating system and controller.
- Drive count, models, capacities and the position of each member.
- RAID type if known, pool/volume layout and SSD cache details.
- Exact errors, log times and order of failures.
- Encryption, prior replacements, rebuilds and configuration changes.
- Backups, snapshots and the last verified good copy.
Keep labels and records with any drives a specialist asks you to send. Do not shuffle members between bays or let different people make unrecorded changes.
What to avoid while the situation is uncertain
Do not create a new pool, initialise disks, force members online or run repair/rebuild commands simply to see what happens. A rebuild is a write operation within a specific recovery plan, not a read-only file-recovery scan. Follow model-specific instructions only after establishing that they fit your actual array state.
For shared production systems, coordinate preservation with your IT team. Agree how to stop further writes and whether controlled shutdown or another containment step is appropriate; do not abruptly change a live system without understanding its dependencies.
Backups and snapshots: check the actual restore point
Ask when the required data last existed intact. Verify that the restore point contains that version and that you can access it independently of the failed storage. RAID redundancy is not a separate backup, and a snapshot on the affected system may not be accessible after its failure.
If a usable backup exists, restoration onto healthy storage may be the quickest route. Keep the original system intact until the restored files and applications have been checked.
Choose a provider for your system
Ask about the exact RAID/controller, file system, encryption and any virtual machines or databases. A provider should describe what it needs for assessment and what parts of your case remain uncertain. Do not infer expertise from a broad “all RAID supported” claim.
Request a plan covering imaging or preservation, analysis, priority data, delivery and validation. For an application, specify whether you need files, a mountable volume or a working database checked by your own team.
Business priorities, permissions and confidentiality
Name one incident lead and one person who can approve costs. Rank the files or services needed first, record the deadline, and agree who will validate them. Ask about subcontracting, file access, secure transfer and retention. A higher emergency fee is only useful if its promised delivery fits your needs.
Software may be used by experienced practitioners for suitable cases, but this guide does not supply a universal rebuild recipe. Use our provider checklist and cost comparison questions before committing.
Frequently asked questions
Can I rebuild first and recover later?
Not as a general rule. Establish the actual fault and preserve what is needed before authorising changes that write to the array.
Does RAID mean I have a backup?
No. Redundancy and a separate usable copy are different protections. Check the actual backup and restore point.
Is NAS recovery only for large businesses?
No. Home and small-office NAS devices need the same care about array state and copies, even when fewer people use them.
Sources and methodology
Synology: causes of volume crashes; Ontrack: preservation and RAID rebuild cautions, checked 2 October 2026. Incident and enquiry checklists are our editorial recommendations, not model-specific repair instructions.
Read our editorial policy and affiliate disclosure for how we evaluate guidance, distinguish research from hands-on testing and explain commercial relationships.
