I have lost faith!

M

metril

Limp Gawd
Joined
Jun 3, 2007
Messages
414
No, this post isn't about religion. Rather, it is about one of the largest hard drive manufacturers and their line of NAS ceritified drives. I am talking about Western Digital.

Long story short, I have 4 WD RED Drives. Two 3 TB, one 2 TB, and one 1 TB. Three were bought from Amazon between January and June of this year. Only one 3 TB was bought from the Egg sometime in April.

Well, two 3 TBs have failed. SMART shows it as a mechanical failure. When my Linux system warned that the file system had gone to read only, the SMART log showed a "Device fault; Error ABRT". Well, it happened a few times. SMART tests passed. And then the drive died. I've already RMA'd one and the other one died last night. UNRAID started mounting my cache drive as read only and gave me the "sense code" that corresponds to hardware failure. I was able to get to the SMART log once, but now the drive doesn't even respond. I've already tried it with a different controller, cable, computer, etc. This RMA is in process.

The remaining 2 TB and 1 TB drives are also reporting the same "Device Fault; Error ABRT" in the SMART log. No reallocated sectors. No seek errors. Just that same "Device Fault; Error ABRT".

I've lost trust in Western Digital. I used to shy away from Seagate, but my Seagate drives have been strong for a lot longer than these WD drives.

And just now, my 1 TB WD Green drive from late 2011 has also died.


I have tried different cables, different controllers, computers, etc. The drives are at fault.

UPDATE: One PSU is Thermaltake and the other is Antec.
 
Last edited:
Could you be more clear about when the drives were in which computers? Just making sure that you aren't having the drive get damaged in one PC and then attempt to see if it still works in other PCs.

What PSUs do you have in each of these PCs? Do you get power outages or brownouts? Are any of these PCs protected by UPSes?
 
It's hard to believe that many drive failures in the same environment is coincidence, I would start looking at the environment, bad power, controllers, something other then random bad luck.I've had very good luck with WD RED drives, I have 14x3TB in my environment for almost a year now, no issues, but that doesn't mean they won't fail, they do and they will, in your case that's too much bad luck to be just that IMO.
 
As evil and ash have said, while it is always possible that you have that many drives that have failed because of defects in materials or workmanship, it is highly improbable. We have about 250 3TB Reds in production servers right now, and have about a 4% annual failure rate. While that is higher than I would like, it is not out of line. How are the drives mounted? Do you have any vibration mitigation? What MB/Controller/PSU? Straight power from the wall or UPS? What else is on this particular power circuit (AC, refrigerator, heater etc?) Temps? On 24/7 or a lot of Start/Stop cycles due to power off or sleep?
 
I thought the red drives were notorious for being crap, or is it just the ones from newegg? Don't buy the red drives.

I wish samsung was still making drives out of every brand they are the ones that seem to work the best for the longest. 8 years now in my server, and although a couple drives have had a hard time starting up once or twice they keep working.
 
You could give the Toshiba MG03ACA series a spin or the desktop ACA series?
People including myself have reported very good experience with former HGST and Toshiba HDDs.
I've also heard that the RED drives are not so good...
//Danne
 
hows the power at your house? do your lights dim when things kick on? do you have a UPS?
 
I have a UPS. The power does not fluctuate when my AC turns on or I flick on other appliances.

The drives were not swapped between systems. I only swapped them between systems to check if it was due to some other device/cable/etc.

I have an Antec PSU in one system and a OCZ in the other. The PSUs are 550W or better.

If it was a PSU issue, then I would expect the other drives on the same rail to fail too. But, the other drives on the same rail are fine.
 
That is odd that two of your drives have failed and two others are having issues. Seems like an issue with power delivery as others have mentioned but it sounds like you have that in hand. :confused:
 
You leave them on all the time? External HDDs get hot quickly if you do that due to poor cooling and heat can easily kill a HDD. I only turn my external HDDs on when I am actually backing up or retrieving data from them.
 
They are internal drives. The cases that the drives are in has 1 120mm intake fan per 3 drives in the cage. There is 1 rear exhaust fan, 1 top exhaust fan, the PSU exhaust fan, 1 exhaust fan behind the motherboard, and 1 300mm intake fan on the the side panel. The fans do not distribute any noticeable vibration since they are all attached using rubber grommets. The cases are quite heavy.

The NAS drives are meant to be able run 24/7. My drives temps are around 30-35 C. I've rarely seen them go above 33 C.
 
Okay. Here's an update folks.

For the first 3TB drive that I RMA'd, I received a recertified drive in return. I ran badblocks on it and let it go through the default 4 passes. I use unRAID as my storage OS. So, I ran one instance of preclear. That passed. Then I decided to be sure and started another instance. I come back from work today and I'm wondering why the preclear shows 41 hours when the first run only took 31 hours. Matter of fact, it wasn't done yet. So I stopped it and that's when I noticed that the recertified drive had dropped out. It was undetected.

I was suspicious of the Thermaltake power supply so I bought an EVGA NEX750G from Microcenter. So, it's highly possible that 2 drives failed just due to failing and 2 drives failed due to a bad power supply. The Thermaltake power supply is 1.5+ years old. It's a 500W.

Running preclear on the recertified drive again. WE SHALL SEE!
 
I never have to use any UPS for my computers at home.
I've never used OCZ PSU, so I have no comment on it.
Antec PSU killed 2 mobos in the past from what I remember. It has real good cases but stay away from its PSUs.
All my home PCs/ servers are now running on SeaSonic or Cosair PSUs only.
You have what you pay for.
 
For the uneducated, not all Seasonic PSUs are great but they're at least good. The OCZ ZT-range are good PSUs too.
Corsair have some crap PSUs so going be brand in general is a bad idea unfortunately. Actually a few of the Rosewill PSUs are awesome despite the brand(ing) =)
//Danne
 
Last edited:
I've never had any problems with Antec PSU's, I have a True Power Trio in my fiances rig and my brother has had one running strong for 7 years, though if I remember right they were built with SeaSonic parts :D. Just have to research and make sure not to buy anything cheap.
 
To some degree I think it depends on a few things:

The quality of your power, the quality of the parts in the power supply and their ability to handle good, not so good or bad power, the components in your rig and how long you run your rig for.

Personally, I learned long ago that dropping anywhere from $80 for an OEM to over $200 retail for a quality ps was well worth it, no question.
 
You must log in or register to reply here.
Back
Top