The three major HDD manufacturers are selling slower drives, without telling us

[Vishera]

2 minutes ago, MyName13 said:

So that's why my HDD is slower than my ancient Maxtor...

What you quoted is not accurate,some P300 are SMR and some are CMR,

The P300 500GB,1TB,2TB and the 3TB models are all CMR,the rest of the P300 models are SMR.

[Curufinwe_wins]

4 hours ago, leadeater said:

I'd say basically already. Couldn't really see anyone recommending it for that right now. But even then it looks like problems exist outside of FreeNAS so I'd be hesitant to recommend until a firmware update for the disks comes out or someone can show exactly how/why the fault condition comes up.

 

Problem is Ironwolf cost more and is a performance tier up from WD Red, cost is higher than Red Pro even.

 

I miss the days before WD killed it when HGST was a strong midrange/budget cost, high performance NAS option. So angry with WD recently.

 

I mean if you can feel confident that the WD drive is CMR, it would be fine, but good luck with getting that info since WD is being absolute asshats about the situation still. [At less than 8TB drives, the 8-14 TB drives are all CM helium drives and should be fine. Expensive, but fine. For now...]

[leadeater]

13 minutes ago, Vishera said:

What you quoted is not accurate,some P300 are SMR and some are CMR,

The P300 500GB,1TB,2TB and the 3TB models are all CMR,the rest of the P300 models are SMR.

All these small disks don't make any sense to me being SMR either. To me that belongs in the larger disks, 8TB+, maybe even 10TB+. The size of the CMR zone is based on the number of platters and the more of those the easier it is to assign more capacity to CMR which allows the disks to handle a higher workload without going to in limp mode.

 

Like a 10TB disk with 1TB CMR would be so incredibly unlikely to fill that it will 99.9% of the time act like CMR disk and a fast one at that. Also the dual actuator also makes sense to me, you could split the workload and function of that so you fill one CMR zone then switch to the other zone which is on the other actuator and allow the first to cleanup, flipping back and forward which should prevent the entire drive write problem for ZFS and other RAID rebuilds entirely.

[Curufinwe_wins]

7 minutes ago, leadeater said:

All these small disks don't make any sense to me being SMR either. To me that belongs in the larger disks, 8TB+, maybe even 10TB+. The size of the CMR zone is based on the number of platters and the more of those the easier it is to assign more capacity to CMR which allows the disks to handle a high workload without going to in limp mode.

 

Like a 10TB disk with 1TB CMR would be so incredibly unlikely to fill that it will 99.9% of the time act like CMR disk and a fast one at that. Also the dual actuator also makes sense to me, you could split the workload and function of that so you fill one CMR zone then switch to the other zone which is on the other actuator and allow the first to cleanup, flipping back and forward which should prevent the entire drive write problem for ZFS and other RAID rebuilds entirely.

I think the big drives are sufficient profit and high-end customers that it is being held off by those factors. [Helium drives already command a premium.]

 

PLUS WD has the audacity to claim that the AX SMR drives (in their documentations) have better performance than their same capacity RX (CMR) counterparts. And don't specify read speeds. Like screw you so much WD, I hope you get sued like crazy over this.

[Vishera]

3 minutes ago, leadeater said:

All these small disks don't make any sense to me being SMR either. To me that belongs in the larger disks, 8TB+, maybe even 10TB+. The size of the CMR zone is based on the number of platters and the more of those the easier it is to assign more capacity to CMR which allows the disks to handle a high workload without going to in limp mode.

What i said was from Toshiba's full statement in which they listed all of the CMR drives in their desktop consumer lineup.

[leadeater]

1 minute ago, Curufinwe_wins said:

I think the big drives are sufficient profit and high-end customers that it is being held off by those factors. [Helium drives already command a premium.]

True but it's ass backwards and those don't buy WD Red anyway which have above 10TB capacity disks in that line, they don't even buy Red Pro either. These are all the Seagate Enterprise Capacity customers (Exos E & X) type customers.

 

Plus everything we buy is custom OEM design stuff provided by Seagate, Hitachi, HGST/WD that it doesn't really matter what they do to their own line of products as they still have to meet the spec of HPE or Netapp etc that they agreed to in the supply contract.

 

And when it comes to HPE they list so much more information about the disk and it's performance, you know exactly what you're getting e.g.:

https://h20195.www2.hpe.com/v2/getpdf.aspx/a00001287enw.pdf

 

Problem is the way it works for enterprise OEMs is your pricing is determined by how much you order per year and how much they value you as a customer, so I could be getting the same disk at half the price someone else is.

[leadeater]

7 minutes ago, Vishera said:

What i said was from Toshiba's full statement in which they listed all of the CMR drives in their desktop consumer lineup.

Yea I got that, was just stating it doesn't make sense to me to do it on the bottom half of the capacity end instead of the top half. Seems like there is more cost saving to be gained there than in the lower end. I dunno just weird. More just a statement overall, since that's what WD has done. Edit: And I guess Toshiba shares that sentiment.

[Kisai]

WD and Seagate are now pulling the "well then use the right drive, idiot" type of response.

 

https://arstechnica.com/information-technology/2020/04/seagate-says-network-attached-storage-and-smr-dont-mix/

 

Quote

Seagate confirms that we do not utilize Shingled Magnetic Recording technology (SMR) in any IronWolf or IronWolf Pro drives—purpose-built for NAS solutions. Seagate always recommends to use the right drive for the right application.

 

Quote

This sounds almost identical to Western Digital's justification for surprising NAS users with SMR drives in the Red line—a statement that the slower disks are "appropriate for the application." Belloni acknowledged the similarity but reiterated Seagate's position on IronWolf and NAS disks—and pointed out that both its Exos and Archive SMR disks have always been clearly marked as such.

https://blog.westerndigital.com/wd-red-nas-drives/

https://blog.westerndigital.com/wp-content/uploads/2020/04/2020_04_22_WD_SMR_SKUs_1Slide.pdf

 

 

[Vishera]

Just now, Kisai said:

WD and Seagate are now pulling the "well then use the right drive, idiot" type of response.

 

https://arstechnica.com/information-technology/2020/04/seagate-says-network-attached-storage-and-smr-dont-mix/

 

 

https://blog.westerndigital.com/wd-red-nas-drives/

https://blog.westerndigital.com/wp-content/uploads/2020/04/2020_04_22_WD_SMR_SKUs_1Slide.pdf

 

 

In Seagate's case it's fine,only Western Digital is in the wrong here.

[leadeater]

2 minutes ago, Kisai said:

WD and Seagate are now pulling the "well then use the right drive, idiot" type of response.

Anyone putting a non NAS rated disk in to a RAID array actually has no basis to complain or mount a consumer complaint so Seagate has nothing to answer for at all, they weren't the one putting SMR in to a NAS rated and advertised disk unlike WD who did and did a bad job at it on top of that.

 

Seagate's statement is completely fair, buy a NAS disk and it'll be exclusively CMR from us.

[MyName13]

37 minutes ago, Vishera said:

What you quoted is not accurate,some P300 are SMR and some are CMR,

The P300 500GB,1TB,2TB and the 3TB models are all CMR,the rest of the P300 models are SMR.

Well, it's still incredibly slow (or maybe Windows 10 simply hates HDDs).

[Curufinwe_wins]

8 minutes ago, Kisai said:

WD and Seagate are now pulling the "well then use the right drive, idiot" type of response.

 

https://arstechnica.com/information-technology/2020/04/seagate-says-network-attached-storage-and-smr-dont-mix/

 

 

https://blog.westerndigital.com/wd-red-nas-drives/

https://blog.westerndigital.com/wp-content/uploads/2020/04/2020_04_22_WD_SMR_SKUs_1Slide.pdf

 

 

 

Could you imagine getting a new laptop with a 1TB internal HDD and it turns out to be SMR because WD is being actual asshats and many laptop makers don't actually care themselves?

[Yes they use oem-specific drives, but based on this positioning... Like WTF a SMR drive gets the WD black label? Are you flipping kidding me? Reminder that the Black line is supposed to be their highest performance consumer line.]

[leadeater]

1 minute ago, Curufinwe_wins said:

Like WTF a SMR drive gets the WD black label?

Lol what? How? Why? Can't wait for 10K RPM HDDs with SMR.

[Curufinwe_wins]

Just now, leadeater said:

Lol what? How? Why? Can't wait for 10K RPM HDDs with SMR.

It's flipping insane and BS. Bury them alive for this crap. I'm very unhappy that Seagate swapped drive techs without telling anyone. If they didn't know it was shittier, they wouldn't try to hide the swap. But at least in their case, as you've said, NAS drives are unaffected. WD's positioning is just inexcusable.

[Vishera]

4 minutes ago, MyName13 said:

Well, it's still incredibly slow (or maybe Windows 10 simply hates HDDs).

When a hard drive is in 100% usage it will be crawling to a slow.

Just now, leadeater said:

Lol what? How? Why? Can't wait for 10K RPM HDDs with SMR.

That specific model @Curufinwe_wins pointed at is actually a 7200 RPM rip-off drive.

[Vishera]

4 minutes ago, Curufinwe_wins said:

I'm very unhappy that Seagate swapped drive techs without telling anyone. If they didn't know it was shittier, they wouldn't try to hide the swap.

That shouldn't concern you and honestly doesn't matter at all for consumers,The read and write speeds are still true to what is written in the datasheets and reviews.

[Curufinwe_wins]

4 minutes ago, Vishera said:

That shouldn't concern you and honestly doesn't matter at all for consumers,The read and write speeds are still true to what is written in the datasheets.

That is fucking bullshit, how dare you suggest that a company without notice replacing products with inferior variants shouldn't concern me or consumers in general.

 

One: WD doesn't advertise write and read speeds (they advertise a single rated 'maximum transfer speed'). Two: when the drives fail at their explicitly advertised purpose, that isn't excusable whatsoever.

 

WD rates for yearly load, I can be well within 180 TB/year and still massively run afoul of a DMSMR drive's ability to handle the situation.

[leadeater]

4 minutes ago, Curufinwe_wins said:

I'm very unhappy that Seagate swapped drive techs without telling anyone. If they didn't know it was shittier, they wouldn't try to hide the swap.

I can understand in the desktop market where large sustained writes aren't that common so experience shouldn't change but there is more than one way to do it. You could do it on the Barracuda line, drop the price of that and then drop the price of Barracuda Pro to the previous Barracuda pricing. Or create a new line like Barracuda S or Flex or Lite or EVO w/e and have that a price point down.

 

Problem with every way no matter what all the OEMs will go for the cheaper one so a makes practically no difference to what you'll get in the majority of products, the name of it become moot really.

 

All OEM computers with HDD boot devices need to die anyway, all of them.

[Sophia_Borjia]

31 minutes ago, Kisai said:

WD and Seagate are now pulling the "well then use the right drive, idiot" type of response.

Seagate can say that, drives fit use advertised for.  

 

WD saying don't use wrong drive, after advertising wrong drive as correct and changing specs that effect what product line is correct for a use case with no indication....not so much.

 

If WD had just put smr on product pages when they made change they could still say that.  SMR Red would still work for 2 bay nas, jbod, mirrors, ect.  People would just know not to use it with raid types/software that doesn't play with smr well and buy different model for that. 

[Curufinwe_wins]

Here is WD's specifications for two 6TB WD Red drives. AX is the SMR variant (which has only been discovered in the last few weeks). Other than the suspiciously (but not unprecedented) large cache, please tell me how any user is possibly supposed to know or realize that one of these drives is basically always going to fail a zfs resilver for example.

 

[Curufinwe_wins]

2 minutes ago, leadeater said:

I can understand in the desktop market where large sustained writes aren't that common so experience shouldn't change but there is more than one way to do it. You could do it on the Barracuda line, drop the price of that and then drop the price of Barracuda Pro to the previous Barracuda pricing. Or create a new line like Barracuda S or Flex or Lite or EVO w/e and have that a price point down.

 

Problem with every way no matter what all the OEMs will go for the cheaper one so a makes practically no difference to what you'll get in the majority of products, the name of it become moot really.

 

All OEM computers with HDD boot devices need to die anyway, all of them.

Creating a new line, or specifically pointing out within the same line (with appropriate disclaimers) would be acceptable.

 

And yes they do, but if a CMR boot device is bad... a SMR boot device is damn near unusable.

[Vishera]

Just now, Sophia_Borjia said:

 

Seagate can say that, drives fit use advertised for.  

 

WD saying don't use wrong drive, after advertising wrong drive as correct and changing specs that effect what product line is correct for a use case with no indication....not so much.

 

If WD had just put smr on product pages when they made change they could still say that.  SMR Red would still work for 2 bay nas, jbod, mirrors, ect.  People would just know not to use it with raid types/software that doesn't play with smr well and buy different model for that. 

You quoted the wrong person...I have never said what you quoted.

[leadeater]

2 minutes ago, Curufinwe_wins said:

Here is WD's specifications for two 6TB WD Red drives. AX is the SMR variant (which has only been discovered in the last few weeks).

 

                                                                                                            ^^^^ Hmmm..... Already knew about this but just pointing it out, I also use that to figure out which disks are SMR now that weren't before.

[Curufinwe_wins]

1 minute ago, leadeater said:

                                                                                                            ^^^^ Hmmm..... Already knew about this but just pointing it out, I also use that to figure out which disks are SMR now that weren't before.

That they don't report weight? Why is that done (actual question here)? Fewer platters weighing less? Also for a non-professional (and also some pros I bet unfortunately), if they did report a lower weight, how many would realize that that meant it must be using SMR and thus be avoided?

[leadeater]

2 minutes ago, Curufinwe_wins said:

And yes they do, but if a CMR boot device is bad... a SMR boot device is damn near unusable.

It shouldn't be, SMR disks have a CMR zone which is required for these to even operate at all. The larger the disk the larger this zone is. This zone is used for all writes, they land there first, get sorted/rearranged and the affected tracks in the SMR zone also copied here then the data written back to the SMR zone.

 

As long as you don't overwork the CMR zone it is a CMR disk. This is why it doesn't belong on small disks. On an 8TB Seagate Archive disk the CMR zone is 25.6GB so as long as you don't copy greater than 25.6GB all at once performance is the same as a CMR disk. 

 

So for something like a 1TB disk the CMR zone is only going to be something like 3.2GB or less. SMR does not belong on these small devices.