HDD error message ever since installing a non identical disk ?

[xlr8r]

hey guys

i have been receiving emailed error messages on a disk ever since it was installed (5-6 years ago). The disk is almost identical to the other 5 (4TB) Seagate discs.

i.e.

5 x ST4000VN000-1H4168

1 x ST4000DM000-1F2168

strangely coincidentally, the error produced is on the "odd man out" DM000 disk.. ;

i.e.

ReadyNas Alert Message (NAS Disk:Critical)

Detected high uncorrectable error count:[56] on disk 6 (internal)

[ST4000DM000-1F2168] This condition often indicated an impending failure. Be prepared to replace this disk.......

I have always assumed this to be a false positive considering it has done it from day one, since setting up a clean OS and 6 new disks. The NAS has been working perfectly regardless of the disk error for a number of years now (5-6). But doubt is always at the back of my mind every time i see the error in my emails...

How can i be 100% sure disk 6 is really failing ? (given that it has been in use without issue for past 5-6 years ? (note, i did the OS6 upgrade  (currently on latest , 6.10.5) to my Readynas RNDP6000 about 3 years ago, and it still produces the same disk error on that same disk. 

Could the error be just because the disk is not identical to the other 5 ...?

hoping someone can advise...thanks

disk log and smart log attached below.

disk_info 

smart_history

[StephenB · Accepted solution]

---

@xlr8r wrote:

is it safe to do so even though I am not getting any "degraded raid messages" on the NAS.

---

Yes, you can just hot-swap the disk.  

Netgear does recommend making sure your backup is up-to-date first, as the array isn't protected during the resync.  So errors with one of the existing disks during that process will result in data loss.

---

@xlr8r wrote:

(does new drive need formatted first?)

---

No.  And if it is pre-formatted, you will need to format in the NAS before it will add it to the array.

Personally I do test the disks before I add them (using Seatools in the case of Seagate disks).  I run the full generic test, followed by the full erase test (which is a bit tricky to get to with Seatools).  I'd strongly recommend doing that with a recertified drive.

[StephenB]

---

@xlr8r wrote:

I have always assumed this to be a false positive considering it has done it from day one, since setting up a clean OS and 6 new disks. The NAS has been working perfectly regardless of the disk error for a number of years now (5-6).

It's a bad idea to ignore disk errors.

It's not because you are mixing models.  In this case, there are definitely SMART errors being being detected:

Command Timeouts: 15
Current Pending Sector Count: 64
Uncorrectable Sector Count: 64

While these might pass a factory diagnostic (not sure if they will or not), I'd just replace it.

Though mixing models is ok, many users have reported issues with the DM model (4 TB Barracuda) in their ReadyNAS. The VN (Ironwolf) model is a better choice.

---

@xlr8r wrote:

How can i be 100% sure disk 6 is really failing ? (given that it has been in use without issue for past 5-6 years ? (note, i did the OS6 upgrade to my Readynas RNDP6000 about 3 years ago, it still produces the same disk error on that same disk. 

---

RAID might be correcting the underlying problem (which is linked to a failed read).

In my experience, you really don't want to wait until you are 100% sure, as that means waiting until the disk has actually failed.

You could run the disk test on the volume settings wheel, and see what happens with that.  But a new 4 TB Ironwolf is fairly inexpensive (around $100 USD in the US), so I'd just replace it and move on.

[xlr8r]

thanks StephenB

i had actually ran the disk test from the cog wheel in Volumes but i expected it to appear in the logs, but unless ive missed it, it doesnt ?

I do recall seeing a popup screen showing it was doing the tests , but i just closed it thinking it was doing it in the background...? Should the window have been kept open for the duration of the tests ?

i did see the LCD display messages that the tests were in progress and then completed. Should a log have been produced ?

please advise thanks 

ps.The VN (Ironwolf) model in the UK is at least twice the US price.... 😞

[StephenB]

---

@xlr8r wrote:

I do recall seeing a popup screen showing it was doing the tests , but i just closed it thinking it was doing it in the background...? Should the window have been kept open for the duration of the tests ?

It's fine to close the window.  There should be a completion message in the logs.  If the test fails, it will say so, if there is just a completion then it passed.  The test does take a while, and since it is a background operation it will take longer if the NAS is in active use.  

One aspect here is that different people have different views about when a disk is "failing".  There is no real standard here.  I replace disks when I see cmd counts and reallocated+pending sectors start to suddently rise, or when the reallocated+pending sector count reaches the 20s. I'll also replace them if I see recurring errors, even if the SMART stats aren't rising.  Others will use different thresholds (some stricter, others more relaxed).

Another is that tests do miss stuff.  Before I install a new drive, I test it by running a full non-destructive test, followed by a full erase test.  I'll confirm problems when I take a disk out of service using the same tests.  With Seagate disks, I use Seatools for these tests. Sometimes the erase test will fail when the non-destructive test passes - and I've also seen the opposite (where the erase passes but the non-destructive test fails).

If you have SSH enabled, you could try 

smartctl -x /dev/sdf

That will give you a lot more information than you'll find in the logs.  That includes an error log, which will give you some idea about when the errors occured.  That can help in understanding if the disk is stable or not.  Note that this error log is kept by the disk itself - that is why the times are given in power-on hours.  

[xlr8r]

cheers 

based on your earlier post, i gave Seagate a call on their online chat, and a very helpful lady there decided to RMA the disk even though way out of warranty!!

it wont be an exact match replacement, nor could she offer the VN000, but said it would be a new 4TB. I suppose what i could do it now buy a VN000 and resell the drive seagate send me....

meantime, i will also check your suggestions using SSH for more info

many thanks

[xlr8r]

StephenB

a question on replacing the failing drive, with a new drive. (RAID 5)

is it safe to do so even though I am not getting any "degraded raid messages" on the NAS. I know this can have adverse effects doing so on the Synology NAS, does same rule apply to Netgear NAS ?

....or do i simply yank out the offending drive from my nas while powered up and replace it , and wait ? (does new drive need formatted first?)

thanks

[StephenB · Accepted solution]

---

@xlr8r wrote:

is it safe to do so even though I am not getting any "degraded raid messages" on the NAS.

---

Yes, you can just hot-swap the disk.  

Netgear does recommend making sure your backup is up-to-date first, as the array isn't protected during the resync.  So errors with one of the existing disks during that process will result in data loss.

---

@xlr8r wrote:

(does new drive need formatted first?)

---

No.  And if it is pre-formatted, you will need to format in the NAS before it will add it to the array.

Personally I do test the disks before I add them (using Seatools in the case of Seagate disks).  I run the full generic test, followed by the full erase test (which is a bit tricky to get to with Seatools).  I'd strongly recommend doing that with a recertified drive.