News WD Sets the Record Straight: Lists All Drives That Use Slower SMR Tech

Toggle sidebar Toggle sidebar
Meanwhile I have two new 6TB Red Drives that I use in my UNRAID NAS (one as a parity) that I have to question if they will fail when they fill up. I guess I can pull them and return them but I have to order two new drives to replace them and that will be a mess. Or I'll have to buy a new Parity drive just to be on the safe side.

Sure as hell won't be WD. I've been reading on reddit how people are having all kinds of problems with the ZFS Raid. This mess has drained them of my good will. I paid a premium for those drives over other brands ($150/drive) I'm sure an OS patch could fix it, if the OS realizes it's a SMR drive. The OS would have to trim, wait and retry after a couple minutes. That might work if you are rebuilding an array. But if you are dumping in new data from an outside source, that doesn't work so well. Try telling windows backup to wait 3 minutes before it can write another file.
 
Last edited:
  • Like
Reactions: martinch
Yea think I'll be buying Seagate now. They are even a little cheaper.

The thing I don't understand is that SMR is suppose to help increase density on large drives. I would think the 8+ TB Drives would be using this NOT 1-6TB Drives.
 
drtweak said:
Yea think I'll be buying Seagate now. They are even a little cheaper.

The thing I don't understand is that SMR is suppose to help increase density on large drives. I would think the 8+ TB Drives would be using this NOT 1-6TB Drives.
Click to expand...

It's because any given amount of storage they sell is going to be cheaper to make with SMR as opposed to CMR, regardless if it's 1TB or 8TB, because less material is needed to manufacture it.
 
  • Like
Reactions: bit_user
digitalgriffin said:
Meanwhile I have two new 6TB Red Drives that I use in my UNRAID NAS (one as a parity) that I have to question if they will fail when they fill up. I guess I can pull them and return them but I have to order two new drives to replace them and that will be a mess. Or I'll have to buy a new Parity drive just to be on the safe side.
Click to expand...

Check the model number and cache of your drives. I have four 6TB Reds in my NAS, and they're CMR.
WD60EFAX - 256 MB Cache - SMR version
WD60EFRX - 64 MB Cache - CMR version

drtweak said:
Yea think I'll be buying Seagate now. They are even a little cheaper.
Click to expand...
Seagate was doing the same thing: https://www.tomshardware.com/news/sneaky-marketing-toshiba-seagate-wd-smr-drives-without-disclosure
 
  • Like
Reactions: bit_user and martinch
digitalgriffin said:
Meanwhile I have two new 6TB Red Drives that I use in my UNRAID NAS (one as a parity) that I have to question if they will fail when they fill up. I guess I can pull them and return them but I have to order two new drives to replace them and that will be a mess. Or I'll have to buy a new Parity drive just to be on the safe side.

Sure as hell won't be WD. I've been reading on reddit how people are having all kinds of problems with the ZFS Raid. This mess has drained them of my good will. I paid a premium for those drives over other brands ($150/drive) I'm sure an OS patch could fix it, if the OS realizes it's a SMR drive. The OS would have to trim, wait and retry after a couple minutes. That might work if you are rebuilding an array. But if you are dumping in new data from an outside source, that doesn't work so well. Try telling windows backup to wait 3 minutes before it can write another file.
Click to expand...
Swore off of WD ages ago - aren't the WD Reds for small 8 drive NAS? - similar to the Seagate Iron Wolf - with the Iron Wolf Pro being for larger arrays and the Exos being for the largest arrays.

Pretty crappy for Seagate and WD to not tell people
 
  • Like
Reactions: digitalgriffin
Lucky - back when I built my NAS I decided 8 tb was the sweet spot. It looks like they decided to use SMR on their big volume drives, and that was 6 tb and below... I'm hesitant to fill up my NAS now, if by chance they decide to implement DMSMR on 8 tb too...
 
Right now all (Seagate, WD and Toshiba) of the major manufacturers have been selling SMR without clearly branding them as such.

And is really worrying. SMR have such a huge drop in writing performance than they had to advertise them as "archiving" devices before. Until all of them come clear and clearly brand SMR devices as such, i will try not to buy an HDD.
 
  • Like
Reactions: martinch
Vorador2 said:
Right now all (Seagate, WD and Toshiba) of the major manufacturers have been selling SMR without clearly branding them as such.

And is really worrying. SMR have such a huge drop in writing performance than they had to advertise them as "archiving" devices before. Until all of them come clear and clearly brand SMR devices as such, i will try not to buy an HDD.
Click to expand...
Actually that's not entirely true, SMR drives have pretty decent write speeds but in limited scenarios - they are great at sequential transfers, but random kills them.

I am surprised that their newer high capacity drives do not exclusively use SMR and they've filtered it down to the lower tiers - SMR seemed a great way to reach those new capacity limits without actually needing to spend as much effort properly increasing the data density / track spacing on the media itself - now they are using it to economise.

I mean a WD Black SMR drive.... WTF?? They'd have been better off just making the black product line SSD or SSHD only without SMR. The blue series I can semi-understand but as they will likely sometimes still be boot drives that's still a random write scenario that doesn't play to SMR's behaviour very well.
 
  • Like
Reactions: bit_user
The problem with the WD RED SMR drives in RAID isn't that they're slow (although the performance hit is 90-98% at times(*)).

The problem is that they have a FIRMWARE BUG which causes them to generate bogus fatal IO errors during raid rebuilds(resilvering) and when running verification passes(scrubbing)

For obvious reasons, the computer/RAID controller then decides it has a bad drive

(*) How bad is it? Well how about 12-20 hour RAID rebuilds now taking 8-10 DAYS? - on a prosumer drive range where there's been no change to the package or pricing and the things have worked perfectly well for the last decade.

WD's gaslighting its customers on this (and hiding the changes, and explicitly refusing to state what changes they'd made when people complained) after making unannounced changes that tank performance is the kind of thing that makes magistrates presiding on cases go "exemplary damages"
 
  • Like
Reactions: artk2219 and martinch
  • Like
Reactions: bit_user
So if its not listed its not SMR? They didnt list their entire product line? What about their Gold drives? In my case they didnt make black drives large enough, so my PCs have Gold drives in them. Didnt find anything on their site that references any of this either?
 
stoatwblr said:
The problem with the WD RED SMR drives in RAID isn't that they're slow (although the performance hit is 90-98% at times(*)).

The problem is that they have a FIRMWARE BUG which causes them to generate bogus fatal IO errors during raid rebuilds(resilvering) and when running verification passes(scrubbing)

For obvious reasons, the computer/RAID controller then decides it has a bad drive

(*) How bad is it? Well how about 12-20 hour RAID rebuilds now taking 8-10 DAYS? - on a prosumer drive range where there's been no change to the package or pricing and the things have worked perfectly well for the last decade.

WD's gaslighting its customers on this (and hiding the changes, and explicitly refusing to state what changes they'd made when people complained) after making unannounced changes that tank performance is the kind of thing that makes magistrates presiding on cases go "exemplary damages"
Click to expand...
Not to defend WD or other drive makers but many NAS implementations seem to ignore the ever increasing likelihood of SMR drives being in a pool. The drive makers and NAS manufacturers (as well as those responsible for software based drive pooling like ZFS, Storage Spaces, etc) should have got together to agree proper information the drive could post about its features such as SMR block sizes, non-SMR cached areas, etc., on drives where the SMR zones are drive-managed so that the software can better manage RAID rebuilds and do calculations of block sizes the drive can just commit without causing overlap and needing to read and re-write any blocks, causing further IO hang-ups.
Or just allow the user to flag its an SMR drive and the software takes a cautions approach and only processes huge data chunks (say 256MB per drive - a value I know Seagate have used as SMR block size) when it does rebuilds or other RAID maintenance - the only problem with that is a huge amount of RAM that would be required on the host and no doubt more CPU intensive.

Instead the approach was just to recommend not to use them... "problem solved, knock off early guys."
 
  • Like
Reactions: bit_user
drtweak said:
Yea think I'll be buying Seagate now. They are even a little cheaper.

The thing I don't understand is that SMR is suppose to help increase density on large drives. I would think the 8+ TB Drives would be using this NOT 1-6TB Drives.
Click to expand...

Seagate is doing the same thing.

And the last new seagate drive I bought, a 2.5" 'firecuda' had half its rated performance, ran hot and had more media errors than drives I've had for ten years.

To answer your other question, there's no money in small drives. SMR lets you cut costs (and performance) on the lower profit margin items.

What I'm not able to conflate is a WD Black drive using slow tech and WD saying "Oh, its an okay tech for every day use". I'm not buying a performance Black drive for "everyday use".
 
  • Like
Reactions: bit_user
I wonder if the preclear utility will show the weakness.

preclear_disk.sh - a new utility to burn-in and pre-clear disks for quick add

A utility to "burn-in" a new disk, before adding it to your array has been requested several times. Also requested is a process to "pre-clear" a hard disk before adding it to your array. When a special "signature" is detected, the lengthy "clearing" step otherwise performed by unRAID is skipped. ...
forums.unraid.net forums.unraid.net
 
This issue is just a peripheral point.
The real issue is that you do not know the geometry of the HDD that you are purchasing.
If you download the PDF file with the specs, you get some generic information. And at the bottom it says:
"Specifications are subject to change without notice"
The reason you do not know the geometry of the WD HDD you are buying is that WD plans to give you some random HDD from some random manufacturing plant in China that could be almost anything that a random manufacturer happens to have in their inventory at some random point in time.

WD has no idea what HDDs they will have available next week.
 
This whole thing frustrates the ever-living crap out of me. I recently upgraded my NAS drive that is used for raid backup and as a server for our home computers to access and share files. I have a WD My Cloud that I purchased years ago and upon recent inspection the old 2 TB drives are CMR. But we were running out of space so we upgraded to 2 6TB drives. It has been feeling very slow since that upgrade. Then this news came to light. I researched a fair bit before I bought them, and if I had this info I would have gone with the pro version, which is only about $50 more per drive. I guess I am now looking to change drives again. My Plex server will have some added storage if I ever get around to building it.
 
Mr5oh said:
So if its not listed its not SMR? They didnt list their entire product line? What about their Gold drives? In my case they didnt make black drives large enough, so my PCs have Gold drives in them. Didnt find anything on their site that references any of this either?
Click to expand...
With the Gold drives, I think they will be very clear about whether they use SMR. Gold are intended for enterprise applications that potentially do continual random I/O that wouldn't give DM-SMR a chance to coalesce the random writes that it has buffered.

Furthermore, Gold are probably most often used with hardware RAID controllers that will timeout a drive extremely quickly. For such drives that are intended for RAID usage, they have a feature called TLER (Time-Limited Error Recovery) that will limit how long the drive tries to re-read data from a block with parity errors. So, they definitely understand that some drives have serious responsiveness requirements.

I used Gold drives in my (software) RAID, primarily due to the warranty term (5 years) and reliability claims (2 M hour MTBF). Sometimes, the uncorrectable read error rates are also an order of magnitude lower, but I think not in my case.
 
@PaulAlcorn , thanks for covering this. It seems they've learned a lesson I hope they don't soon forget.

Honestly, most people still using HDDs are probably using them in RAIDs. Sheesh.

BTW, back when I had a NAS, the manufacturer maintained a list of recommended/qualified HDD models. People should be encouraged to check whether their NAS vendor does this, and to buy disks from that list.
 
bit_user said:
Furthermore, Gold are probably most often used with hardware RAID controllers that will timeout a drive extremely quickly. For such drives that are intended for RAID usage, they have a feature called TLER (Time-Limited Error Recovery) that will limit how long the drive tries to re-read data from a block with parity errors. So, they definitely understand that some drives have serious responsiveness requirements.
Click to expand...

To a certain extent, wasn't that the point of the Black drives also (long warranty, suitable for - but not targeted at - RAID usage, better overall product with the best internals)?
WD progressively hobbled the range from having TLER, probably to ensure sales of their raptor/gold drives as the ultimate product, to the point you'd almost only purchase the black for the extra warranty and slightly better numbers over the blue range (especially now in the age of the SSD it's almost just a choice of "you can have one of our hard drives OR a slightly better one, but still nowhere near as fast as an SSD, for a bit more but with a longer warranty also").
Now some have SMR in their 'performance' product... Again I'm not saying SMR is slow for transfers but random write is the enemy.
Hell, their first SSHD was in the WD Black family - that's the kind of new features people want to see, not SMR.

Can't help but think this is a by product of dropping the green line into the blue so effectively no longer having a 'budget' version of the red and hence ending up with the Red Pro as the 'real Red' drives and the current red as just the budget low energy always on drive.
Looking at it like this makes it more logical but this is where consolidation of branding works against you. WD are having the latest version of a silent spec change causing a PR moment like Kingston when they changed NAND spec mid product life cycle.
 
Last edited:
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom