Synology RAID failure and why RAID is not a backup
I wanted to share a situation I recently dealt with that reminded me why RAID is not a backup.
A Synology NAS device (DS1512+) that I manage notified me that one of the drives in the array had fallen out:
Volume 1 () on ********** has entered degraded mode. (Total number of hard disks: 5; Number of active hard disks: 4) Please identify the cause of degradation and take corrective action.
To repair the volume, please replace the defected disks with new ones. For more information, visit Synology’s Knowledge Base on our official website.
No big deal, that’s what RAID is for. I logged into the unit and confirmed that indeed: Disk 1 – Crashed
I had a few spare hard drives available, so I quickly swapped the drive and the RAID began rebuilding from its degraded state. All was well with the world, another example of RAID doing it’s job of ensuring data reliability!
Until 45 minutes into the rebuild I got an additional message:
Volume 1 () on ********** has crashed. The system may not boot up.
To obtain further assistance, please contact Synology Online Support: http://www.synology.com.
I logged in and confirmed – during the RAID rebuild Drive 4 had also crashed – rendering the entire volume lost.
A few minutes of research reveals several reasons why RAID is not backup.
Things that RAID does not protect against:
- Corrupt file
- Accidentally deleted file
- System compromise (virus/malware)
- Catastrophic disaster (fire/tornado/earthquake)
At the end of the day RAID protects against one type of failure: Physical media failure
In my case, that’s exactly what it did. It protected my data volume against a single drive failure.
Problem was, all the drives were about the same age, and had experienced the same level of usage. A RAID rebuild is one the the most taxing things you can do to a hard-drive and sure enough, a second drive died during the rebuild process.
Fortunately I wasn’t completely reliant on RAID as my primary means of ensuring data reliability. This Synology unit was being backed up to another NAS unit.
Due to two consecutive drive failures (drives were nearing 7 years age) – I replaced all 5 drive bays, and restored the data once the RAID volume had been re-created.
Morale of the story: RAID is not a backup! Ensure your critical data has a secondary means of performing a recovery. If you don’t have a secondary Synology unit – most Synology units interface with cloud-based storage and Amazon Glacier is really cheap!