NAS TROUBLE GUIDE
Multiple NAS drives
show failures or warnings
When several members show warnings, the priority changes from repairing the NAS to preserving the evidence needed to reconstruct the array. RAID fault tolerance describes how many failures a layout may survive under ideal conditions; it does not guarantee that every remaining sector is readable during a rebuild.
Do not treat every warning as the same event
Two members show Warning
Both may still be readable, or one warning may be secondary. Capture SMART summaries, I/O errors and timestamps without launching extended tests.
One Failed, another Warning
A rebuild will depend heavily on the warning member. Do not assume it can survive a full pass merely because files are still accessible.
Two or more Missing/Failed
The layout may be beyond normal repair. Do not create a new array, initialize members or shuffle bays to see whether it boots.
Array still appears online
Online access does not prove redundancy is intact. Minimize writes and verify an independent backup before any repair workload.
Checks that do not alter the evidence
- Stop new writes and heavy jobsPause users, sync, recording, backup destinations, scrubs and media indexing when this can be done without destabilizing the unit.
- Capture the array and pool screensRecord RAID level, pool state, volume state, each member status, error text, timestamps, serial numbers and firmware version.
- Preserve bay order physicallyPhotograph the front of the chassis and drive labels. Do not rely on memory or assume identical drives are interchangeable.
- Verify the failure sequenceWrite down which warning appeared first, any power event, drive replacement, reboot, firmware update or rebuild attempt.
- Verify a copy outside this arrayOpen critical files from another device or cloud location. A snapshot stored in the affected pool is not a separate backup.
When to continue and when to stop
Do not do this
- Replacing several members in one operation
- Testing every original member in a normal PC and accepting an initialization prompt
- Changing bay order or discarding the first drive marked failed
- Starting a rebuild while another member reports I/O errors or mechanical noise
- Creating a new array, pool or filesystem over the original members
- Writing recovered files back to any source member
Verify against official documentation
- QNAP QTS 5.2: RAID group statusOfficial meanings of Ready, Degraded, Rebuilding, Read-only and Not Active.
- Synology DSM 7: RAID types and fault toleranceGeneral RAID characteristics and supported failure counts.
- QNAP: Rebuilding a degraded RAID groupThe supported workflow after confirming that routine repair remains appropriate.
- Synology: Repair a storage poolReplacement requirements and the official DSM repair path.
Menu names, supported repairs, drive requirements and hot-swap behavior vary by model, firmware and RAID layout. Use the current documentation for your exact device before changing disks or array settings.
Frequently asked questions
Can I replace two failed drives one at a time in RAID 6?
RAID 6 can generally tolerate two failed members, but warning, failed and unreadable are not interchangeable states. A rebuild must read all surviving members, and another read error can stop it. Confirm the exact array status, bay order, backup and manufacturer procedure before replacing any member.
Should I remove each drive to test it in a PC?
Do not do that as a first step. A PC may prompt you to initialize an unfamiliar filesystem, and removing members can change the NAS state or lose the bay order. Record the array and drive information in place. If forensic imaging is needed, it should be planned with write protection and correct member ordering.
Is a Warning drive the same as a failed drive?
No. Warning may reflect SMART attributes, I/O events, temperature or other conditions while the drive is still participating in the array. Record the exact message and do not assume that the first warning drive is the only problem.
PRESERVE THE ORIGINAL STATE
Record first. Change nothing you cannot undo.
Photograph the status screen, bay order, drive labels and exact error message before seeking model-specific support or a professional evaluation.
See all English safety guides →Scope: multi-bay NAS and RAID appliances · Last reviewed: August 28, 2026