@masukomi
SMART warning means the drive is running out of "spare" sectors to allocate in the event of failed sectors in the future - no actual current problems, but it is losing the ability to compensate for regular wear and tear.
Fortunately you have it in RAID. Depending on exact configuration, it should be able to recover from one or more entire drive failures - which this isn't.
I'd replace the warning drives one at a time, and give it time to rebuild the data from parity while operating
@masukomi
One of your earlier screenshots showed a bad sector count that would have me considering replacement. Especially if that number keeps going up. Failures tend to be fairly consistently low for the driver's life and spike towards the end.
If it was my system I'd immediately replace the drive with the highest bad sector count and check the other one regularly. RAID can survive one drive failure, so I'd probably pinch pennies and try not to replace both but YMMV for peace of mind.
@masukomi
In any case, RAID in your most likely configuration should only have one drive replaced at once, whether that is precautionary or failure, because it needs all the other drives to reconstruct the data for the new drive.
@Smingleigh Thank you for the advice. I REALLY appreciate it. The bad sector numbers were disconnected from any real meaning for me beyond the obvious fact that some sectors were bad which by itself isn't too notable.
My understanding is that it'll survive 1 drive failure, and then reconfigure itself to survive another, with less available space in the interim. BUT I don't wanna go there ;)
@masukomi
Just to be clear, once a drive in the array has failed, it cannot recover from future failures unless the failed drive is replaced and given time to reconstruct the data before the next failure. It will only ever tolerate one drive failure - your 3-drive redundant array will not become a 2-drive redundant array on failure, it will be fully dependent on its redundancy capacity just to keep operating, probably at worse performance.
@Smingleigh Hmm. Apparently I made bad assumptions about how "Synology Hybrid Raid" functions. They don't really describe it in too many details though. :/
@masukomi@Smingleigh As suggested backup the data and start replacing the drives one at a time. The risk of another failing drive during rebuild is high as the stress on the array is intensive. So having multiple drives flagged should be taken seriously.
SHR vs conventional raid is only that it allows a combination of disks that are not of the same capacity. Meaning it will utilize them better but in some cases that makes sense with more drives (number of disk and their size).
Add comment