Register for the iXsystems Community to get an ad-free experience and exclusive discounts in our eBay Store.

Warning: WD Red WD60EFRX-68L0BN1

tobiasbp

Member
Joined
Dec 2, 2015
Messages
238
It seems there is a reproducible problem with WD Red disks of this (exact) type WD60EFRX-68L0BN1. In my 24 disk zpool, those drives always throws errors during scrubs.

Maybe the 1-8 disk rating for WD Reds is not just marketing. On the other hand, an older model of the same drive (WD60EFRX-68MYMN1) does not have the problem.

The model WD60EFRX-68L0BN1 seems to be the latest model as of fall/summer 2017.

Please read my original thread for more details.

Do you use WD60EFRX-68L0BN1 drives? What is your experience? How many drives is in your zpool?

Thanks,
Tobias
 

Dice

Neophyte Sage
Joined
Dec 11, 2015
Messages
1,215
Hey.
Interesting to hear.
I got a mixed bag of said drives running in the same vdev as RaidZ2:
Code:
root@Lagoona:~ # smartctl -a /dev/da0 | grep "Device Model"
Device Model:	 WDC WD60EFRX-68MYMN1
root@Lagoona:~ # smartctl -a /dev/da1 | grep "Device Model"
Device Model:	 WDC WD60EFRX-68MYMN1
root@Lagoona:~ # smartctl -a /dev/da2 | grep "Device Model"
Device Model:	 WDC WD60EFRX-68MYMN0
root@Lagoona:~ # smartctl -a /dev/da3 | grep "Device Model"
Device Model:	 WDC WD60EFRX-68MYMN1
root@Lagoona:~ # smartctl -a /dev/da4 | grep "Device Model"
Device Model:	 WDC WD60EFRX-68L0BN1
root@Lagoona:~ # smartctl -a /dev/da5 | grep "Device Model"
Device Model:	 WDC WD60EFRX-68L0BN1
root@Lagoona:~ # smartctl -a /dev/da6 | grep "Device Model"
Device Model:	 WDC WD60EFRX-68L0BN1


yet I've never seen any scrub related issues.
 

tobiasbp

Member
Joined
Dec 2, 2015
Messages
238
yet I've never seen any scrub related issues.
It looks as if you have 7 drives in your zpool. If so, the drives really should work with no issues (Since they are rated for raids up to 8 drives).

If you
 

Dice

Neophyte Sage
Joined
Dec 11, 2015
Messages
1,215
I recon that is nothing but a sales-argument to keep consumers to get REDs for 'consumer grade NAS' while they aim the WD PRO and other higher series for 'pro environments'.
The drive don't know there are other drives next to it.
The <only> argument would be from a vibration standpoint. Yet - that is a question about mounting and may very well be far more flimsy in a consumer grade NAS than a SC846 or SC847 for example.

I believe the 'rant about max number of drives according to manufacturer' has been brought up in the forums before.
Sales BS, AFAIK.
 

BlueMagician

Member
Joined
Apr 24, 2015
Messages
55
Just to add my reply in this newly created thread as well as the original...

I am running a 6 disc WD60EFRX (Red) RAIDz2, comprised of three different revisions of these drives.

During almost every scrub, an error email is generated, reporting that a SCSI MEDIUM error (URE) occurred - always caused by the same drive - the only 68L0BN1 in my pool.

I have tried swapping the drive to a different HBA port, and the error follows the drive, so it is unlikely to be a controller or data cable issue. All drives are running the same firmware, and all report 'healthy' in SMART stats and tests...
 
Last edited:

Chris Moore

Wizened Sage
Joined
May 2, 2015
Messages
10,066
I believe the 'rant about max number of drives according to manufacturer' has been brought up in the forums before.
Sales BS, AFAIK.
I agree, I use consumer grade desktop drives in my NAS and have no problems with them. I would be more inclined to think there is some other problem with the system than that there is a fault in the drives.
Do you use WD60EFRX-68L0BN1 drives? What is your experience? How many drives is in your zpool?
What are the other components involved in the connection to the drive, system board, HBA?
 

Johnnie Black

Neophyte Sage
Joined
May 10, 2017
Messages
834
It does indeed look like an issue with those disks, though maybe some hardware combination triggers the problem, controller would be the #1 suspect but @Dice appears to also be using an LSI and he's unaffected, strange problem.
 

tobiasbp

Member
Joined
Dec 2, 2015
Messages
238
@Dice appears to also be using an LSI and he's unaffected, strange problem.
I think the number of drives in the enclosure is a factor. I should put a "bad" drive back in my pool and confirm errors appearing. Then I should hang the drive off of the machine using cables (Vibration proof). If the errors disappear, I would think vibrations are the course.
 
Last edited:

Chris Moore

Wizened Sage
Joined
May 2, 2015
Messages
10,066
I think the number of drives in the enclosure is a factor. I should put a "bad" drive back in my pool and confirm errors appearing. Then I should hang the drive off of the machine using cables (Vibration proof). If the errors disappear, I would think vibrations are the course.
Have you contacted Western Digital about this? If you have some drives of the same model that work correctly and others that are giving errors, I would say it is simply a hardware fault in the drive and the drive needs to go in for warranty replacement. If you contact WD, don't tell them how many drives you have in the NAS. They will happily say, "not my problem," if they can and leave it at your feet.
 

tobiasbp

Member
Joined
Dec 2, 2015
Messages
238
If you have some drives of the same model that work correctly and others that are giving errors, I would say it is simply a hardware fault in the drive and the drive needs to go in for warranty replacement.
That's not the case.

Drives of this type always throws errors (I have 6):
WD60EFRX-68L0BN1

Drives of this type never trows errors (I have 7+):
WD60EFRX-68MYMN1
 

Chris Moore

Wizened Sage
Joined
May 2, 2015
Messages
10,066
That's not the case.

Drives of this type always throws errors (I have 6):
WD60EFRX-68L0BN1

Drives of this type never trows errors (I have 7+):
WD60EFRX-68MYMN1
If you are getting errors, there is defective hardware. It wouldn't be the first time a manufacturer put out a product that was bad from the start. To me it sounds like you need to open a support incident with Western Digital because they have produced a defective product. The number of drives should not cause the drive to throw errors. Admittedly, the drives I have knowledge of are not the same model, but I have many Western Digital drives under my watchful eye. One of the server chassis at work has 60 Western Digital drives in it and I don't get errors there and I have two more 24 bay chassis filled with Red Pro drives that don't give me errors and two more 24 bay chassis filled with WD Red (not pro) drives and I get no errors there. Then there are all the Seagate and HGST drives in other chassis that don't give me errors... If you have errors, there is a fault in the system.
 

smitelli

Newbie
Joined
Nov 30, 2017
Messages
1
Jumping in with my experience here. I'm in the process of upgrading my 8-disk pool from 3TB drives to 6TB, and all the drives I've been receiving are WD60EFRX-68L0BN1.

After installing the sixth drive (so, pool had 6x WD60EFRX-68L0BN1 and 2x old WD30EFRX-68AX9N0) a scrub occurred. About 45% of the way through, I get a single "... SCSI sense: MEDIUM ERROR asc:11,0 (Unrecovered read error) ..." event in the syslog and the scrub status starts "repairing" the disk named in the error message. When the scrub finishes, about 120K of data is repaired in total. This is reproducible -- repeating the scrub causes the error to recur, different disk (but always a 68L0BN1) and different adapter port. No SMART errors, and as far as I can tell no corruption or degradation of the pool data.

8 disk RAID-Z2 on an IBM ServeRAID M1015 / LSI SAS 9240-8I (IT mode).
 
Last edited:

tobiasbp

Member
Joined
Dec 2, 2015
Messages
238
When the scrub finishes, about 120K of data is repaired in total. This is reproducible -- repeating the scrub causes the error to recur, different disk (but always a 68L0BN1) and different adapter port.
This is exactly what I was seeing too. My theory is, that the 68L0BN1 is a cost reduced (inferior) version of the drive, that behaves correctly when fever drives are used together (Matching WD's product description). I this scenario, the older drives are probably exceeding their specs.

In my 24 disk pool, I have since seen the occasional SCSI error on other WD60EFRX (Non 68L0BN1) drives.

I am replacing all my WD60EFRX drives with WD Golds. I have never seen an error on any of those.
 

Chris Moore

Wizened Sage
Joined
May 2, 2015
Messages
10,066
How many of those are of the type 68L0BN1?
All the drives I have in the servers at work were bought retail as bare drives.
I can look at the details when I get to work.

Sent from my SAMSUNG-SGH-I537 using Tapatalk
 

tobiasbp

Member
Joined
Dec 2, 2015
Messages
238
For the record: I have not had problems after I replaced all WD Red disks with WD Golds.
 

Chris Moore

Wizened Sage
Joined
May 2, 2015
Messages
10,066
The 6TB drives that I have at work are the Red Pro drive in a 60 drive pool. The WD Red (non pro) drives are the 4TB model, so my experience with the drives is not necessarily applicable to your situation. Still, the 4TB WD Red drives in the system at work have no given us a problem and we have 64 of them in that system with the rest of the pool being Seagate Constellation drives.

Sent from my SAMSUNG-SGH-I537 using Tapatalk
 

tobiasbp

Member
Joined
Dec 2, 2015
Messages
238
I'm not sure I understand this correctly....
The 6TB drives that I have at work are the Red Pro drive in a 60 drive pool.
You have a pool of 60 6TB WD Red Pro drives?


The WD Red (non pro) drives are the 4TB model, so my experience with the drives is not necessarily applicable to your situation. Still, the 4TB WD Red drives in the system at work have no given us a problem and we have 64 of them in that system with the rest of the pool being Seagate Constellation drives.
You have another pool with 64 4TB WD Red's and an unspecified number of Seagate drives?
 
Top