Synology NAS suddenly fails to load, storage pool crashes? An explanation of the reasons and self-help boundaries

I. Four Common Postures of NAS Crash

Synology is widely used in small and medium enterprises and design industries, and failures are also common. According to our repair statistics, there are roughly four categories:

• System partition corruption: Can ping but DSM web page cannot be boarded, or prompt "system partition failure", SSH to see the system disk mount abnormal. Multi-cause power failure, system upgrade failure, memory instability.
• Storage pool/volume degraded or crashedIn DSM, the Storage Manager reports "Storage Pool Crashed","Volume Corrupted", and one or more disks display "Uninitialized","System Partition Failed".
• Hard drive mass outageBackplane power supply or SATA channel problems cause multiple disks to fall at the same time, and the disk itself may be good.
• human error: Shared folder deleted by mistake, volume deleted by mistake, disk with data initialized, RAID type migration failed.

Let's make it cleardisk problemstillsystem problems, the processing path is completely different.

II. Case: Design company DS920 +, system partition damage after a power failure

On August 18, 2026, a brand design company in Nancun, Panyu, had an office building trip when rushing to finish manuscripts at night. After a call came in, DS920 +(4 4TB Western Digital red disks, SHRarray, containing customer source files, material library and company accounting photos) About 9TB) The machine could not be turned on: the power light turned on, the status light was amber, the DSM webpage turned on timeout, and the display reported "System partition failure, please reinstall DSM".

The administrative girl almost clicked the prompt "reinstall." Here's a key sentence:The Synology prompt "Reinstall DSM" usually moves only system partitions without touching data partitions, but "Initialize""Rebuild Storage Pool" clears data.If you can't tell, don't light it. Cut the power and send it for repair. They did the right thing. They brought the machine.

Treatment process:

1. Measure the disk first, do not move the array.Dismantle four disks and connect them to read-only equipment one by one to detect SMART and surface: Disk 3 has 47 remapped sectors (05 counts) and is still growing, and the remaining three are healthy. Judgment: Tripping and power failure damaged Disk 3. After SHR degradation, the system repeatedly retried under abnormal state, and the system partition collapsed.
2. Full-disk mirroring.All four disks are read-only mirrors. Disk 3 uses slow +bad sectors skip mode and mirrors for 9 hours.
3. Linux environment reorganization.Synology SHR is essentially mdadm soft RAID (this is SHR=RAID5 variant)+LVM+ext4. In the recovery environment, mdarray is reorganized by mirror, LVM volume is activated, ext4--directory tree is completely ejected, customer source files and material libraries are all in place. ext4's journal mechanism helps a lot in this power failure, and file system consistency is very small.
4. Export and relocation.9TB data exported to the customer's newly purchased NAS (this time it is recommended that they install DS923 + with 4 pieces of 8TB, and enable snapshot function). Sampling: PSD/AI source files randomly opened 50, all normal, Lightroom directory reconstruction index photos complete.

Fees: NASarray reorganization and recovery (4 disks of SHR/9TB of valid data) is RMB 3200, and bad sectors disk images are added for RMB 400, totaling RMB 3600. The data export uses a new NAS purchased by the customer, and the media fee is confiscated.

3. Which situations can be saved by themselves and which must be stopped

Can try it on your own (low data risk):

• Only DSM web page cannot be boarded but SSH can enter: DSM service may be hung up, restart NAS (press the machine power button to shut down normally and then open) Most can be good.
• Single disk reports "uninitialized" but storage pool is still degraded available status:Copy out the key data while array is online, and then consider changing the disk repair. Degradation status is the last window period, do not take the window period to do reconstruction.
• Error delete shared folder and open recycle bin: go to #recycle directory to find.
• Snapshot function (Btrfs volume +Snapshot Replication) enabled: rollback directly from snapshot, which is the best regret medicine for Synology. It is recommended that all Btrfs users enable it, and the space occupied can be controlled.

What must be stopped:

• Storage pool crashed + multi-disk exception: any "repair storage pool""initialize hard drive" button do not click.
• There are abnormal noises on the disk, SMART reports a lot of bad sectors: first mirror and then talk about something else, directly tossing on the original disk is equal to secondary damage.
• Exception after RAID migration/expansion interruption: array metadata may be in an intermediate state, and it is highly probable that it will mess up the disk order.

IV. Technical cards for data recovery (why can they be saved)

Qunhui's storage structure is public: Linux mdadm soft RAID (SHR/RAID1/5/6/10)+ LVM 2 + ext4 or Btrfs file system. This means that as long as there is a full-disk image, you can reorganize the array and mount the file system in the standard Linux environment (or professional tools such as PC-3000 and UFS Explorer), independent of whether the original machine can be started or not. This is also the reason why the recovery success rate of NAS is generally not low-it is much more friendly than the encrypted master of SSD.

There are two things that are really hard: one isBtrfs metadata extensively corrupted(Ext4 is relative to skin), and the second isDisk order/RAID parameters are artificially scrambled and then written. So the old adage goes back: After an accident, write operations are the biggest enemy.

5. NAS data recovery fee reference in 2026

Similarly, recovery is unsuccessful without recovery fees, diagnosis and assessment free of charge.

Six, four life-saving suggestions for NAS users

1. Equipped with UPS.If there is a UPS of 500 or RMB 600, the accident will not happen at all. NAS is most afraid of abnormal power failure. A medium-end UPS (such as TG-BOX 850VA) can last for 15 minutes to make NAS gracefully shut down.
2. Take a snapshot.Btrfs Volume +Snapshot Replication, ransomware and deletion regret medicine, set up a tube for several years.
3. NAS is not the only backup.The NAS itself can be stolen, burned, and blackmailed. Critical data can be added to another offsite (cloud or another physically isolated hard drive).
4. Keep an eye on SMART.DSM storage manager can see the hard drive health, 05 (remapping sector), C5 (to be mapped), C6 (irreparable) three counts are non-zero on the spare disk, do not wait for it to drop.

NAS accident, first power off, do not click any "initialization/repair" button, and then call 020-39029800. Report clearly the model, disk number, RAID type and what you did before the accident, the phone can help you determine the direction.