THE STORY SO FAR
Last Septemebr my MyCloud EX4 had problems and after interacting with WD support they replaced the unit and the 4x4TB drives in ealy October. It ran well till late Feb when it reported Disk 4 failed. This time I decided to go as far as I could without WD support. I ordered in a new red disk, swapped it in and the RAID 5 re-build ran for some time then failed. I retried a number of times and sometimes it complained disk 3 had failed and once or twice dsik1. One odd things with the attempts was that I had done full factory restores which promised to wipe the data, but then after some reboots it saw the 4 TB of data and you could see it in Windows, but then it would fail and the file visibility went.
Anyhow, I ordered another new disk and swapped so now I had the original disks 1 & 2 and new reds in 3 & 4. Then it rebuilt OK everything checks out and after 4 days I almost have all the data back.
So I have these 2 “failed” disks. So I get an Orico 2-bay HDD dock and explore thru Windows 10. On disk 4 I have had Windows Disk Manager create a single partition then chkdsk /f /r /x so it looks for bad sectors - it reports none.
MY FIRST QUESTION
Should I return the “failed” disk to WD for replacement, or do I figure that the MyCloud just got it wrong?
Can anyone suggest more tests I should try?
Any other comments are welcome.
I should note that along the way I have contacted WD support for advice and have a number.
Thanks for all this information. So it sounds like Windows 10 chkdsk may be happy with the drive because it will wait longer when sectors are not so good, whereas MyCloud is more impatient.
I have just run WD Digital Lifeguard Diagnostics on the first disk to fail (#4 using the original numbering) and the quick test passes, but the extended test fails before it gets to 1% saying “too many bad sectors”. So maybe I am getting the answer I am looking for, and should be seeking replacements.
When the troubles were in progress I did look at the SMART data but my untrained eye could not pick up useful clues.
Your final para got me thinking. I have had the 4 disks as RAID 5 yielding 11.8 TB of usable space, but when it gets in trouble and has trouble re-building you are faced with losing it all and rebuilding from backups which is SO SLOW. For now I have the RAID 5 re-built and will let it run for now. But possibly I would be better off with two RAID 1s using 2 disks each if that’s permissible. Then a single disk problem would be less disruptive. So if I hit another problem in the coming months I may switch over to that.
One final thought. These were presumably new drives from WD that gave trouble after I had them only 5 months. I am wondering if there are ever “bad batches” of hard disks and the other two may also be questionale. I guess we will see with time.
Thanks again for your considered reply, I really appreciate it.
The most new Disks have bad sectors that are reallocated. Sometimes they are fully identified only by first use. If the reallocated sector did not increase over the next time, the Disk is OK.
That factory low level format is made under optimal circumstances (eg temperatures) and if the bad sectors after the first hours of use did not increase they are just bad sectors which are not identified during the factory initialization. But there should be only a few bad sectors, otherways the reserve sectors could run out.