Any bad experiences with ironwolf 8tb yet?

[diskdiddler]

With or without helium?

 

[Johnnie Black]

With or without helium?

I believe all of them use helium.

 

[Chris Moore]

WD Red is PMR, Seagate ST8000DM004 is SMR.

That is a really bad assertion to make when you can point to no evidence and there are other manufacturers that make a non shingled drive at the same capacity. You are voicing a personal theory, without substantiation, as a fact.

 

[Johnnie Black]

That is a really bad assertion to make when you can point to no evidence and there are other manufacturers that make a non shingled drive at the same capacity. You are voicing a personal theory, without substantiation, as a fact.

I already proved before that using PMR only it's impossible to achieve the areal density of those disks, but if you need more, Seagate has since launched the v3 Archive drives, just compare the specs from both disks, they use the same platters:

https://www.seagate.com/www-content...a-fam/barracuda-new/en-us/docs/100805918d.pdf
https://www.seagate.com/www-content...seagate-archive-hdd/en-us/docs/100827317b.pdf

 

[Stux]

ST8000DM004 is TGMR

Tunnelling Giant Magneto-Resistive
https://www.theregister.co.uk/2011/11/01/seagate_barracuda_1tb_platters/

Page 12:
https://www.seagate.com/www-content...a-fam/barracuda-new/en-us/docs/100805918g.pdf

But who knows...
https://lime-technology.com/forums/...gate-expansion-external-not-an-archive-drive/

 

[Johnnie Black]

ST8000DM004 is TGMR

Yes, but that is the type of head, it has nothing to do with SMR, the new Archive v3 also uses a TGMR head:

From the linked pdf: ST8000AS0003

Code:

Affordable efficiencies with 2TB-per-disk Drive Managed SMR-based hard drive technology

Recording method TGMR

I find the fact that 1100Gb/Sq.in. is the max possible theoretical areal density with PMR only, with current max being below 1000GB/sq.in., is proof enough that a drive with 1200GB/Sq.in. areal density must be SMR, add to that the new v3 Archive using the exact same platters and confirming it's SMR, don't know what more proof is needed.

 

[MrToddsFriends]

[...] don't know what more proof is needed.

Telltale benchmark results would be nice to see, for example comparable to those on storagereview.com for a Seagate Archive HDD Review (8TB).

 

[southwow]

Sorry to have started a religious war. Thanks to everyone who continues to contribute. I've ordered a few 12TB ones from newegg and will report back.

Spent the weekend sick, so not much progress other than resurrecting another system to FTP files to in the interim.

I've also come to the conclusion that I hate bhyve completely, totally, and utterly. What's happened to VirtualBox compatibility?

I'll enumerate the reasons bhyve is bad:
1.) EFI shell won't boot legitimate OS ISO images, then the same ISO works at other times
2.) Formatting a volume is painful
3.) Doesn't play well with > 1 virtualized CPU with any windows OS I've seen yet
4.) What is up with the network stack/bridge? It works sometimes and others it doesn't
5.) Raw File doesn't work in my experience at all
6.) Zero support for host filesystem access

Seriously, this hypervisor feels completely half-baked. It seems like it is designed to prevent it from doing anything useful

 

[Chris Moore]

Sorry to have started a religious war. Thanks to everyone who continues to contribute. I've ordered a few 12TB ones from newegg and will report back.

You didn't ask about 12TB drives. There have been many users reporting problems with the 10 and 12 TB units.

I've also come to the conclusion that I hate bhyve completely, totally, and utterly. What's happened to VirtualBox compatibility?

I agree.

Seriously, this hypervisor feels completely half-baked. It seems like it is designed to prevent it from doing anything useful

I would have thought it was better to improve the PHP Virtual Box that was already available instead of creating something new that doesn't work well.

 

[Ericloewe]

Sorry to have started a religious war.

What's one more? vi vs. emacs is still ongoing.

 

[southwow]

What's one more? vi vs. emacs is still ongoing.

Everyone knows vi is the superior editor. I won't even let my children associate with children of emacs users. Once, I opened it out of curiosity and couldn't figure out how to close it. Finally had to kill the process in another terminal. LOL

Maybe we'll have a clear winner about editors when SCO vs. IBM is finally over in another few decades :p

 

[southwow]

You didn't ask about 12TB drives. There have been many users reporting problems with the 10 and 12 TB units.

I agree.

I would have thought it was better to improve the PHP Virtual Box that was already available instead of creating something new that doesn't work well.

No worries, if they fail, they fail. I also did another thing...

2x 8087 - 8088 rear case brackets
1x RES2SV240NC
1x SMC CSE-PTJBOD-CB2
2x 8088 male to male

This buys me 16 more drives in my old 4U case with backplanes. It also permits me to stop taking the effing cover off of the 846 case every 10 minutes to connect something else...

 

[southwow]

HA! Willcall order cancelled on the 12TB ironwolf. Ordered 3 more red 8TB 5400 for a few $$$ less.

Edit: Having a newegg location near you is extremely dangerous.

 

[c32767a]

ST8000DM004 is TGMR

Tunnelling Giant Magneto-Resistive
https://www.theregister.co.uk/2011/11/01/seagate_barracuda_1tb_platters/

Page 12:
https://www.seagate.com/www-content...a-fam/barracuda-new/en-us/docs/100805918g.pdf

But who knows...
https://lime-technology.com/forums/...gate-expansion-external-not-an-archive-drive/

I have shucked a number of those Seagate external enclosures ($149 at amazon, last I checked) for my home NAS and have gotten 2 ST8000AS0002 and 6 ST8000DM004.

I tested them before I put them in the NAS and the AS0002s act like SMRs (sequential writes are OK, but they hate random IO) while the DM004s behave like regular drives.

I have all 8 drives in a raidZ2 and it's been fine as a video archive/backup for my media server..

 

[Chris Moore]

I have shucked a number of those Seagate external enclosures ($149 at amazon, last I checked) for my home NAS and have gotten 2 ST8000AS0002 and 6 ST8000DM004.

I tested them before I put them in the NAS and the AS0002s act like SMRs (sequential writes are OK, but they hate random IO) while the DM004s behave like regular drives.

Thanks for the information. I think that answers the question that @MrToddsFriends had posed about these drives.

 

[southwow]

I have shucked a number of those Seagate external enclosures ($149 at amazon, last I checked) for my home NAS and have gotten 2 ST8000AS0002 and 6 ST8000DM004.

I tested them before I put them in the NAS and the AS0002s act like SMRs (sequential writes are OK, but they hate random IO) while the DM004s behave like regular drives.

I have all 8 drives in a raidZ2 and it's been fine as a video archive/backup for my media server..

That was exactly my experience with 15 of them (now 14, let's have a moment of silence for that drive)

 

[Johnnie Black]

I tested them before I put them in the NAS and the AS0002s act like SMRs (sequential writes are OK, but they hate random IO) while the DM004s behave like regular drives.

They might act differently due to improved caching, but they are still SMR, no way around that, it's impossible to make a 1.2Tb/sq.in PMR disk.

Basically it's like this:

for 2.5" disks, current max platter size for PMR is 750GB/platter, all 1TB/platter disks like Seagates 3/4/5TB and the new 1/2TB WDxxSPZX are SMR.
for 3.5" disks, current max platter size for PMR is 1.5TB/platter, all 2TB/platter disks are SMR.

This doesn't mean you'll never see larger PMR density, but it's reaching the upper limit of what it's possible, like before the limit off what was possible with LMR was reached, first PMR disks had an areal density of 133Gb/sq.in, we are now around 1TB/sq.in, and it's believed that it will never be possible to go beyond 1.1Tb/sq.in with PMR only, so not likely you'll ever see a 2TB/platter PMR disk, and certainly not possible currently.

 

[diskdiddler]

You didn't ask about 12TB drives. There have been many users reporting problems with the 10 and 12 TB units.

I agree.

I would have thought it was better to improve the PHP Virtual Box that was already available instead of creating something new that doesn't work well.

Why did they switch anyhow? Is VirtualBox not available on FreeBSD or something?

 

[diskdiddler]

They might act differently due to improved caching, but they are still SMR, no way around that, it's impossible to make a 1.2Tb/sq.in PMR disk.

Basically it's like this:

for 2.5" disks, current max platter size for PMR is 750GB/platter, all 1TB/platter disks like Seagates 3/4/5TB and the new 1/2TB WDxxSPZX are SMR.
for 3.5" disks, current max platter size for PMR is 1.5TB/platter, all 2TB/platter disks are SMR.

This doesn't mean you'll never see larger PMR density, but it's reaching the upper limit of what it's possible, like before the limit off what was possible with LMR was reached, first PMR disks had an areal density of 133Gb/sq.in, we are now around 1TB/sq.in, and it's believed that it will never be possible to go beyond 1.1Tb/sq.in with PMR only, so not likely you'll ever see a 2TB/platter PMR disk, and certainly not possible currently.

PMR + Helium = 9 platters, if I recall (isn't that the helium record?)
13.5TB.

I'm not a big helium fan, I prefer disks with no 'odd' features, no SMR, no Helium, just something simple which is (likely) more reliable, but I suspect I'm going to have to change my atittude in the future and deal with it.
SMR is still, the devil and I want nothing to do with it though, that sounds inherently terrible for any NAS, excess reads and writes to all disks / heads, just nah, no thanks.


Isn't there 2 more technologies coming? One being the laser heat one and another one similar, microwave or something?

 

[Johnnie Black]

PMR + Helium = 9 platters, if I recall (isn't that the helium record?)
13.5TB.

Largest currently available PMR drives are 12TB, with 8 x 1.5TB platters, there's a 14TB HGST but it's SMR.