Pending Drive Failure - Need Advice

kmmcd wrote:

Can the WD Red (WD60EFRX) be replaced with a WS Red Plus 6TB drive (WD60EFPX or WD60EFZX)? WD60EFRX Red drives seem to no longer be available. Also, the WD Red drives are SMS whereas the Plus drives are CMR. Does this matter?

The WD60EFRX is CMR (not SMR), and you should only use CMR drives with ReadyNAS.  WD rebranded their Red CMR drives as Red Plus some years back, but the WD60EFRX pre-dates that rebranding.

Red Plus are all CMR, so either the WD60EPRX or the WD60EFZX are ok.  Avoid the WD60EFAX if you see it - it is SMR.

kmmcd wrote:

If the drive fails, do I just need to hot remove the failed drive and hot insert the new drive?

That is the procedure I normally recommend.  That said, it is really important to remove the right drive, and things will go very wrong if you accidentally remove the wrong one and then reinsert it.  So if there is any doubt about the slot then you should power down before removing, and then confirm the serial number.  You can then insert the replacement with the NAS powered down.

Netgear does recommend making sure you have a backup before manipulating disks, and I agree with their advice.  At least back up the irreplacable stuff (family photos/videos, documents) while you are waiting for the replacement to arrive.

kmmcd wrote:

About how long should I expect data recovery to take?
Will my data still be accessible if and after the drive has failed, before I have installed a new replacement drive?

If all goes as usual, then the data remains available through the entire process.  The volume will change to a degraded status as soon as you remove the drive, and it will remain degraded until after you insert the replacement and the RAID resync finishes.  The amount of time it takes depends in part on the size of your array, and in part on how much else the NAS is doing during the resync.  I'd expect around 12-24 hours for a 4x6TB array.

Note that the volume is not protected when the array is degraded.  So if another disk fails during the resync process, the volume will fail and you will lose your data.

kmmcd wrote:

Should I pre-emptively replace the "failing" drive before it actually fails?

It depends in part on how many errors you are seeing.  You might want to download the log zip file, and look at disk_info.log and smart_history.log.  Also you can look for errors in systemd-journal.log - that will show errors that are not in the other two logs.

But it also depends on your risk personal tolerance, and the consequence of data loss.

Personally I replace disks when the reallocated+pending sector count (combined) reaches around 20 or more.  I have also replaced some disks that repeatedly throw other errors.  Usually I am replacing them before the NAS says they've failed.