What if I don't replace failing disk in Raid1

Question

I have a home server on Synology DS716+ running Raid1.

Currently Disk1 S.M.A.R.T. status is Failing and also Bad Sector Count displays 2. So obviously I should replace this hard drive, but I am traveling and will not be home for another 2,5 weeks. My questions are:

1. Can not replacing it immediately cause any further issues and/or data loss?
2. Would it be better to disable this hard drive until I can replace it? Is there a safe way to do it?
3. Any other advice in this situation?

I have Hyper Backup configured to back stuff up to USB and obviously Disk2 is still working. But not sure if this saves me or not.

Many thanks!

Answer 1

A note in general: No one can tell you that really for sure but there are some factors that might have an influence.

1. Of course it can. It's not uncommon for drives to fall shortly after another. You might have luck, or not - no one will be able to say that. However, there are several factors that impact the likelyhood that the second drive will fail also:
   • If they're coming from the same production batch. You could check that probably by reading out their serials.
   • During an array rebuild the stress for the remaining drive is considerably higher than during regular usage.
   • If your drives are specifically designed for being used in a NAS environment/are Enterprise-class drives, the chance for failing is lower (I'm using HGST drives based on reports like https://www.backblaze.com/blog/hard-drive-failure-stats-q2-2017/)
   • The bigger your hdd size is and the more space you're using, the longer a rebuild will take --> the higher the stress is for the drive --> the higher is the chance it might fail.
2. Again, this might happen. If you want to be sure no further corruption happens, you need to prevent any access to the array. So removing the drive from the array in the meantime might help.
3. • I know, this might not be an alternative, but consider something different than a 2-disk-solution, especially if you're using disks in the multiple TB-range. A RAID6-setup could sustain two drive failures before getting critical (performance might be a thing tough).
   • Again, not something for the short term, but a filesystem like ZFS can detect errors at runtime and correct them before they become a bigger issue. Combine that with regular Scrubs and a RAID6-configuration and you're pretty safe. Synology NAS aren't really suited for this however, you might have a look at FreeNAS for example.

Answer 2

1. You can go with a single drive - that's what RAID 1 is for - but if the other drive also fails, the array is offline and data likely lost.
2. No. Note that the drive is shown as failing, not failed. Disabling the drive leaves you with a single drive that could also fail at any time. While the drive isn't too reliable any more, leaving it online is actually better in case the 2nd fails.
3. Replace it as soon as you can manage. I'm assuming that the RAID 1 setup wasn't for fun but to serve a purpose. However elaborate, any RAID setup can fail however, that's why you need a reliable backup.

As I also like to point out for these kinds of problems: make sure you patrol/scrub/scan/test the array on a regular basis. Each sector on a disk is subject to signal degradation - some more, some less. Verifying all of them once a month ensures that weak sectors are detected and repaired in time so you can read back long untouched sectors when they're required for rebuilding. Without scrubbing, rebuilds over large disks fail regularly.

Answer 3

Raid 1 = Mirroring, so there is a full copy of the data on each disk. You should be able to run with just the one drive.