Disk death during resync

[Swoofah]

Ok; so her is is the issue. I have a ReadyNAS Ultra 4 + with 4x2TB Seagate Green disks.

On of the disks (number3) died on me, and since i had to replace it, i figured i might as well replace all of them since it had started to run out of free space.

So i bought some new WD RED 4TB disks and started replacing the one that had died.

The hotswap didn't register, so i rebooted the NAS and the resyncing started. When it came to around the 68% marker something happend. RAIDar reports that the status is "Healthy", but the new disk is marked as a "Spare". Frontview on the other hand reports that "RAID sync finished on volume C. The array is still in degraded mode, however. This can be caused by a disk sync failure or failed disks in a multi-parity disk array." and that DISK1 is dead.

If i reboot the NAS again (with or without the new disk), DISK1 is alive and well again. If the new DISK3 is present resync will start again and the same thing will happen.

So i was wondering if there were something i could do to keep DISK1 "alive" until resyncing is finished? Or is a factory reset my only option?

[mdgm-ntgr]

Do you have a backup?

[Swoofah]

Since I can still access the volume after a reboot, I am backing up the essensials, but i do not have the space to back it all up.

So i can survive a reset, but the ideal scenario would be to not loose data :)

[vandermerwe]

There is almost certainly a problem with disk 1 as you have probably deduced, so it will probably need replacing.  As you can't complete the resync your best ( and only ) option is to back everything up and start over with all 4 of your new disks. 

I would test all of them with wd lifeguard long tests before using them.  If there are any errors, send them back.  

You could verify that disk 1 is a problem by powering down, removing it and testing it.

I think if you just keep trying to replace disk 3 you may end up losing the whole volume if disk 1 fails completely. 

[BaJohn]

vandermewe said

I would test all of them with wd lifeguard long tests before using them.  If there are any errors, send them back.

So you would recommend testing ALL newly bought discs before using them:)

Surely the chances of a dead or dying brand new disc is almost zero?

I've never used wd lifeguard myself, but doesn't this take a long time?

[StephenB]

---

@BaJohn wrote:

Surely the chances of a dead or dying brand new disc is almost zero?

I've never used wd lifeguard myself, but doesn't this take a long time?

---

I've had some DOA disks.  All it takes is mis-handling during shipment.

It does take a while, and sometimes I've skipped it.  But it is a good idea to run the long test (and maybe zero the drive on top of that).

If it fails out-ot-the-box, then you can exchange it for a new one.  If it fails later on, you get a recertified drive.

[vandermerwe]

I do test all new disks:

1. It is usually much easier to send back disks soon after purchase rather than going through a warranty process later, I accept some problems won't be exposed by a test on a new drive.  This is particularily important if you are not going to use the drive immediately. 

2. Problems during a resync are easier to troubleshoot if you have already established that the new drive is definitely perfect.

As for recommending it, yes I do.  Whether you do it really depends on your attitude to risk. I agree it is unlikely but then so are many other events that users assume are so rare that they can be ignored.

It does take a long time, especially if you do it using USB to sata, however direct sata connection can long test a 4 TB drive in about 11 hours. 

,