Applicable Products
Scenario
A disk failure can make a volume degraded or inaccessible. What happens next depends on the RAID configuration. In particular, Single-disk and RAID 0 configurations provide no protection against disk failure. Consider the risk and maintain a separate backup before using either configuration.
The case that prompted this FAQ involved a two-disk RAID 0 group on a TS-251D. After one disk failed, the volume became unavailable and the installed apps disappeared. Replacing the failed disk alone could not restore the RAID 0 volume or its files.
Disk Failure Behavior by Configuration
| Configuration |
Protection against disk failure |
What may happen if a disk fails |
| Single |
None |
The storage on that disk becomes unavailable. There is no redundant copy in the configuration to recover it. |
| RAID 0 |
None |
Data is striped across all member disks. Failure of any member can make the entire RAID group and its volume inaccessible. Replacing the failed disk does not recreate the missing data. |
| JBOD |
None |
Disks are combined without redundancy. Data on the failed disk may be lost, and the combined volume or data on other disks may also become inaccessible, depending on the layout and failure. |
| RAID 1 |
One disk |
The group may continue operating in a degraded state using the remaining mirror. A second disk failure before recovery can make data inaccessible. |
| RAID 5 |
One disk |
The group may continue operating in a degraded state. A further disk failure before the group is restored can make data inaccessible. |
| RAID 6 |
Two disks |
The group may continue operating with one or two failed disks. Failure beyond its tolerance can make data inaccessible. |
| RAID 10 |
One disk per mirrored pair |
The group can tolerate a disk failure in a mirror pair, but not the loss of both disks in the same pair. |
| RAID 50 |
One disk per RAID 5 subgroup |
The group can tolerate one failed disk in each subgroup. More than one failure in a subgroup can make the group inaccessible. |
| RAID 60 |
Two disks per RAID 6 subgroup |
The group can tolerate up to two failed disks in each subgroup. More failures in a subgroup can make the group inaccessible. |
These are general failure tolerances, not guarantees that data will remain accessible or that a rebuild will succeed. Actual symptoms and recovery options depend on the NAS model, firmware, storage layout, disk condition, and whether other errors occur.
What to Do
If data is still accessible
- Back up important data to a separate destination as soon as possible. Do not treat RAID redundancy as a backup.
- Verify that the backup completed and that important files can be opened or restored.
- Check Storage & Snapshots for the affected disk and RAID status. Record the NAS model, firmware, disk bay, status messages, and relevant logs.
- For a redundant RAID group, follow the NAS interface or QNAP Support guidance to replace the confirmed failed disk and monitor the rebuild. Do not remove another disk while the group is degraded.
If the volume is inaccessible or the configuration has no redundancy
- Stop making configuration changes. Do not initialize, format, recreate, or repeatedly rebuild the storage group, and do not replace additional disks.
- Keep the original disks and their bay order unchanged. If the data is important and no usable backup exists, contact QNAP Support or a qualified data recovery provider before attempting recovery. Recovery is not guaranteed.
- If a verified backup is available, replace the failed disk as appropriate, recreate the storage configuration if needed, and restore the data from that backup.
In the TS-251D case, the two-disk RAID 0 volume became Not active after one disk failed. Installing a replacement disk by itself could not restore the volume or its files. The advised recovery path was to recreate the storage configuration and restore from backup; without a backup, avoid further changes and seek professional recovery assistance if the data is important.
Prevention
- Choose a RAID type with suitable disk-failure tolerance for the importance of the data. RAID 1, for example, can tolerate one disk failure in a two-disk mirror.
- Keep regular, versioned backups on a separate system or destination, and test that files can be restored.
- Consider backup coverage for the NAS configuration and applications as well as shared-folder data.
- For RAID 0, Single, or JBOD, plan for the possibility that a disk failure will interrupt access to data; these configurations do not provide disk-failure protection.
Related Links
適用產品
情境
磁碟故障可能導致磁碟區降級或無法存取。接下來的情況取決於 RAID 配置。特別是，單一磁碟和 RAID 0 配置不提供磁碟故障保護。考慮風險並在使用這些配置之前維護單獨的備份。
促使此常見問題的案例涉及 TS-251D 上的兩磁碟 RAID 0 群組。在一個磁碟故障後，磁碟區變得無法使用，已安裝的應用程式消失。僅更換故障磁碟無法恢復 RAID 0 磁碟區或其檔案。
磁碟故障行為依配置而異
| 配置 |
磁碟故障保護 |
磁碟故障可能發生的情況 |
| 單一 |
無 |
該磁碟上的儲存空間變得無法使用。配置中沒有冗餘副本可供恢復。 |
| RAID 0 |
無 |
資料在所有成員磁碟上分條。任何成員故障都可能使整個 RAID 群組及其磁碟區無法存取。更換故障磁碟不會重建遺失的資料。 |
| JBOD |
無 |
磁碟組合在一起但無冗餘。故障磁碟上的資料可能遺失，且其他磁碟上的組合磁碟區或資料也可能因佈局和故障而無法存取。 |
| RAID 1 |
一個磁碟 |
群組可能在降級狀態下繼續執行，使用剩餘的映象。在恢復之前的第二次磁碟故障可能使資料無法存取。 |
| RAID 5 |
一個磁碟 |
群組可能在降級狀態下繼續執行。在群組恢復之前的進一步磁碟故障可能使資料無法存取。 |
| RAID 6 |
兩個磁碟 |
群組可能在一個或兩個磁碟故障下繼續執行。超出其容忍度的故障可能使資料無法存取。 |
| RAID 10 |
每個映象對一個磁碟 |
該群組可以容忍映象對中的一個磁碟故障，但不能容忍同一對中的兩個磁碟損失。 |
| RAID 50 |
每個 RAID 5 子群組一個磁碟 |
該群組可以容忍每個子群組中一個磁碟故障。子群組中超過一個故障可能會使群組無法存取。 |
| RAID 60 |
每個 RAID 6 子群組兩個磁碟 |
該群組可以容忍每個子群組中最多兩個磁碟故障。子群組中更多的故障可能會使群組無法存取。 |
這些是一般的故障容忍度，並不保證資料將保持可存取或重建會成功。實際症狀和恢復選項取決於 NAS 型號、韌體、儲存空間佈局、磁碟狀況，以及是否發生其他錯誤。
應採取的措施
如果資料仍然可存取
- 儘快將重要資料備份到其他目的地。不要將 RAID 冗餘視為備份。
- 確認備份已完成，並且重要檔案可以開啟或還原。
- 檢查儲存與快照總管以了解受影響的磁碟和 RAID 狀態。記錄 NAS 型號、韌體、磁碟槽、狀態訊息和相關日誌。
- 對於冗餘的 RAID 群組，按照 NAS 介面或 QNAP 支援的指導來更換確認故障的磁碟並監控重建。在群組降級時不要移除其他磁碟。
如果磁碟群組無法存取或配置沒有冗餘
- 停止進行配置更改。不要初始化、格式化、重新建立或重複重建儲存空間群組，也不要更換其他磁碟。
- 保持原始磁碟及其插槽順序不變。如果資料重要且沒有可用的備份，請在嘗試復原之前聯絡QNAP 支援或合格的資料復原服務提供者。復原並不保證成功。
- 如果有經過驗證的備份，請適當更換故障磁碟，必要時重新建立儲存空間配置，並從該備份中還原資料。
在 TS-251D 案例中，兩磁碟的 RAID 0 磁碟群組在一個磁碟故障後變成未啟動。單獨安裝替換磁碟無法恢復磁碟群組或其檔案。建議的復原路徑是重新建立儲存空間配置並從備份中還原；如果沒有備份，避免進一步更改，並在資料重要時尋求專業復原協助。
預防
- 選擇適合資料重要性的磁碟故障容忍度的 RAID 型別。例如，RAID 1 可以在兩磁碟映象中容忍一個磁碟故障。
- 在單獨的系統或目的地上保持定期的版本備份，並測試檔案是否可以還原。
- 考慮備份 NAS 配置和應用程式以及共用資料夾資料的覆蓋範圍。
- 對於 RAID 0、單一或 JBOD，計劃磁碟故障可能中斷資料存取的情況；這些配置不提供磁碟故障保護。
相關連結
