HGST 8TB SAS HUH728080AL5200

[UhClem]

... I did end up formatting back to 512 logical sectors and the space went back down. I'm not sure what the exact situation is with this though.

=== START OF INFORMATION SECTION ===
Vendor: HGST
Product: H7280A520SUN8.0T
Revision: PD51
Compliance: SPC-4
User Capacity: 7,865,536,647,168 bytes [7.86 TB]
Logical block size: 512 bytes
Physical block size: 4096 bytes
LU is fully provisioned
...

Thanks very much for your effort. While it's disappointing that it didn't result in 8.0TB @ 512e, at least (for my situation), it doesn't get "stuck" @4Kn. Given this, I'll probably format one drive @4Kn just to play/experiment with for a while (never had a 4Kn drive).

P.S. There's more of these available [as of 23Sep20;3pm et] - @OP link

 

[SirCrest]

These disks appear to be from SUN/Oracle. I wonder if the firmware is automatically applying some sort of internal 520b when formatted to 512e since the user-space is the same as before but the drive doesn't specify the protection it did before. Meanwhile on 4Kn it doesn't bother with it.

 

[eroji]

I pulled the formatted disk and plugged it back, but it still didn't show the increase capacity. I also tried creating a dummy pool with the single disk after. That didn't change the size too. Is there something else I can try?

I used the following command for the formatting

sg_format --format --size=512 --six -v /dev/da10

 

[wildpig1234]

I am using used enterprise 4TB as direct download/storage save to back up. it would be nice to have 8TB. but the cost is still high. the 4TB i use is only 30-35$. When these 8TB enterprisse hdd gets to $60 or less then i can buy them. I usually buy in packs of 10

 

[SpamMeDeeper]

So I finally got around to putting my 8TB drives into my test box to zero test for defect. As promised, sharing some data for those thinking of getting some (if they come back in stock).

These are for sure Sun/Oracle drives. My white labels were all sanded (?) to remove branding. I ordered two different times, 3 in one order, 2 in another for a total of 5. All seem pretty healthy and working. In my server they all show 7.86TB formatted as 512E.

Here are the stats on the drives:

Dec 2015
Start Stop Cycle: 9
Power on Hours: 37380

Aug 2016
Start Stop Cycle: 40
Power on Hours: 31103

Dec 2015
Start Stop Cycle: 8
Power on Hours: 37254

June 2015
Start Stop Cycle: 12
Power on Hours: 7318

Dec 2015
Start Stop Cycle: 9
Power on Hours: 37377

Let me know if there is anything else you'd like me to check. Thanks again to OP for finding this deal.

 

[wildpig1234]

what's the read/write amount?

So I finally got around to putting my 8TB drives into my test box to zero test for defect. As promised, sharing some data for those thinking of getting some (if they come back in stock).

These are for sure Sun/Oracle drives. My white labels were all sanded (?) to remove branding. I ordered two different times, 3 in one order, 2 in another for a total of 5. All seem pretty healthy and working. In my server they all show 7.86TB formatted as 512E.


Let me know if there is anything else you'd like me to check. Thanks again to OP for finding this deal.

What's the read/write amount?

 

[SirCrest]

Well all done migrating to the new disks. My drive's 'smartctl -a' if desired.

Manufactured in week 26 of year 2015
Accumulated start-stop cycles:  12
Accumulated load-unload cycles:  313

Error counter log:
           Errors Corrected by           Total   Correction     Gigabytes    Total
               ECC          rereads/    errors   algorithm      processed    uncorrected
           fast | delayed   rewrites  corrected  invocations   [10^9 bytes]  errors
read:          0        0         0         0     176040       5977.421           0
write:         0        0         0         0    1364022     128887.688           0
verify:        0        0         0         0       7578          0.000           0

Manufactured in week 51 of year 2015
Accumulated start-stop cycles:  17
Accumulated load-unload cycles:  1560

Error counter log:
           Errors Corrected by           Total   Correction     Gigabytes    Total
               ECC          rereads/    errors   algorithm      processed    uncorrected
           fast | delayed   rewrites  corrected  invocations   [10^9 bytes]  errors
read:          0       39         0        39    7371366     150309.296           0
write:         0       62         0        62     841089     123965.939           0
verify:        0        0         0         0     106626          0.000           0

Manufactured in week 51 of year 2015
Accumulated start-stop cycles:  10
Accumulated load-unload cycles:  1552


Error counter log:
           Errors Corrected by           Total   Correction     Gigabytes    Total
               ECC          rereads/    errors   algorithm      processed    uncorrected
           fast | delayed   rewrites  corrected  invocations   [10^9 bytes]  errors
read:          0       29         0        29    7371937     152101.744           0
write:         0        5         0         5     479184     114649.799           0
verify:        0        0         0         0      56600          0.000           0

Manufactured in week 51 of year 2015
Accumulated start-stop cycles:  11
Accumulated load-unload cycles:  1547

Error counter log:
           Errors Corrected by           Total   Correction     Gigabytes    Total
               ECC          rereads/    errors   algorithm      processed    uncorrected
           fast | delayed   rewrites  corrected  invocations   [10^9 bytes]  errors
read:          0        0         0         0    5186598      98493.252           0
write:         0       10         0        10     991276      88677.738           0
verify:        0        0         0         0      92551          0.000           0

Manufactured in week 51 of year 2015
Accumulated start-stop cycles:  8
Accumulated load-unload cycles:  1540

Error counter log:
           Errors Corrected by           Total   Correction     Gigabytes    Total
               ECC          rereads/    errors   algorithm      processed    uncorrected
           fast | delayed   rewrites  corrected  invocations   [10^9 bytes]  errors
read:          0        4         0         4    7285046     154879.286           0
write:         0        2         0         2    1501589     108864.267           0
verify:        0        0         0         0     171915          0.000           0

Manufactured in week 51 of year 2015
Accumulated start-stop cycles:  10
Accumulated load-unload cycles:  1544

Error counter log:
           Errors Corrected by           Total   Correction     Gigabytes    Total
               ECC          rereads/    errors   algorithm      processed    uncorrected
           fast | delayed   rewrites  corrected  invocations   [10^9 bytes]  errors
read:          0        1         0         1    5011479     104911.920           0
write:         0        0         0         0     502271      84959.130           0
verify:        0        0         0         0      27625          0.000           0

Manufactured in week 21 of year 2017
Accumulated start-stop cycles:  49
Accumulated load-unload cycles:  180

Error counter log:
           Errors Corrected by           Total   Correction     Gigabytes    Total
               ECC          rereads/    errors   algorithm      processed    uncorrected
           fast | delayed   rewrites  corrected  invocations   [10^9 bytes]  errors
read:          0        0         0         0      36300      60908.270           0
write:         0        1         0         1     383963      96660.315           0
verify:        0        0         0         0       2715          7.805           0

Manufactured in week 51 of year 2015
Accumulated start-stop cycles:  8
Accumulated load-unload cycles:  1542

Error counter log:
           Errors Corrected by           Total   Correction     Gigabytes    Total
               ECC          rereads/    errors   algorithm      processed    uncorrected
           fast | delayed   rewrites  corrected  invocations   [10^9 bytes]  errors
read:          0        1         0         1    7219733     159783.264           0
write:         0        0         0         0     536599     106770.321           0
verify:        0        0         0         0      29311          0.000           0

Considering what these drives could go through, I think they look fine. I'm on ZFS so they work great. Some look to be in the 160TB read range. And considering they're probably rated for 180TB atleast per year, not bad.

 

[wildpig1234]

Considering what these drives could go through, I think they look fine. I'm on ZFS so they work great. Some look to be in the 160TB read range. And considering they're probably rated for 180TB atleast per year, not bad.

As long as you don't have anything super important on them.

are you using them as a direct to storage/back up to file away immediately once drive is filled or are you using them in a regular manner?

 

[zack$]

236 in stock. Feeling that this might be the last hoorah.

 

[chinesestunna]

236 in stock. Feeling that this might be the last hoorah.

Or.... until 10TBs reach this price haha

 

[SpamMeDeeper]

what's the read/write amount?


What's the read/write amount?

Updated with Read/Writes for the five drives I have:

Dec 2015
Start Stop Cycle: 9
Power on Hours: 37380
Read Processed: 163.4TB
Write Processed: 118.5TB


Aug 2016
Start Stop Cycle: 40
Power on Hours: 31103
Read Processed: 357.6TB
Write Processed: 124.6TB

Dec 2015
Start Stop Cycle: 8
Power on Hours: 37254
Read Processed: 119.2TB
Write Processed: 91.8TB

June 2015
Start Stop Cycle: 12
Power on Hours: 7318
Read Processed: 22.2TB
Write Processed: 149.9TB

Dec 2015
Start Stop Cycle: 9
Power on Hours: 37377
Read Processed: 114.7TB
Write Processed: 96.7TB

Also back in stock!

 

[JoshDi]

This is very tempting.... I probably should go to sleep instead of buying 10 of theses to replace my 3tb WD red array....

 

[Jiaxinxi]

This is SAS drive, WD red are SATA drive. unless your enclosure support SAS, this won't fit it.
Good news, money is saved!

 

[SpamMeDeeper]

I pulled the formatted disk and plugged it back, but it still didn't show the increase capacity. I also tried creating a dummy pool with the single disk after. That didn't change the size too. Is there something else I can try?

I used the following command for the formatting

sg_format --format --size=512 --six -v /dev/da10

Did you ever get 512 to show full 8TB? I'd prefer to stick with 512 so I can move drives around between FreeNAS and other boxes, but will format as 4K if needed. TIA.

 

[eroji]

Did you ever get 512 to show full 8TB? I'd prefer to stick with 512 so I can move drives around between FreeNAS and other boxes, but will format as 4K if needed. TIA.

Nope. None of the steps mentioned works. I just settled with the slightly lower size.

 

[Iaroslav]

Some misfortune maybe, but I've got 3 of 10 drives with defects (thus not even formatting) or not spinning up at all. Waiting for a reply from the seller.
UPD - the seller was kind to accept a return, so the case is clear, the rest of the disks are ok under the regular write load now.
UPD2 for one guy who doesn't believe, just two screens, as the third one was clicking endlessly during startup

 

[mmbtrumpet]

I bought some of these before from a different seller. The first batch I got had a serial number that were 0015* and the second batch were 0017*. The first batch had protection turned on and the second batch did not. The first batch I had to reformat to remove the protection and the use the resize command, the second batch I was able to just run the resize command.

To reformat the drives.
sg_format --format --fmtpinfo=0 /dev/sXX

To resize the drives.
sg_format --resize --count=-1 /dev/sXX

 

[heromode]

Some misfortune maybe, but I've got 3 of 10 drives with defects (thus not even formatting) or not spinning up at all. Waiting for a reply from the seller.
UPD - the seller was kind to accept a return, so the case is clear, the rest of the disks are ok under the regular write load now.

A long story not so short, and to especially quote from above: "or not spinning up at all"

It's time to let the cat out of the bag, and to tear up a bag of beef jerky ( do not eat the silica gel )

It was only by happenstance when browsing a merchant price listing of HGST helium filled drive prices, that i came across the (now) infamous (P3) marking after the HUH* code

P3, i asked myself, never seen that before, and duckduckgoing i went. P3 stands for "Power Disable Feature" it turns out, and lo and behold, all those thousands of returned drives were never defective at all.

My personal favourite from the Western Digital PDF linked after this article, beside the very sexy MOLEX TO SATA CABLE picture, is the following hilarious quote:

" We don’t want customers returning drives that they believe are defective when the problem is with the legacy chassis. A smart customer will be aware of this potential issue, and be prepared to take either version for maximum supply flexibility. "

the joke of course being, the opposite of a 'smart customer' would be a 'stupid customer'.

I tried to refrain myself from making this a widely known issue, because of the obvious advantages a "insider with the knowledge" could have when shopping for helium filled hdd's on ebay, looking for cheap drives being dumped as spare parts.

I have personal experience, a so called 'DOA' HGST HUH***** drive that never spinned up and never showed any signs of life, even when connected to a reasonably modern Corsair RM750 PSU using a standard SATA power connector, does indeed happily spin up using a molex to sata power adapter cable on the same PSU, and shows perfect smartctl stats.

So there you have it, not quite the SMR/CMR level of scandal, but just another sign of the issues caused by the lack of competition in the spinning rust markets. (I do love helium drives though)

Mike

https://documents.westerndigital.co...h-brief-western-digital-power-disable-pin.pdf

 

[kapone]

Needed more storage...so... (already had two of these, so 24 in total).

 

[Sean Ho]

It was only by happenstance when browsing a merchant price listing of HGST helium filled drive prices, that i came across the (now) infamous (P3) marking after the HUH* code

P3, i asked myself, never seen that before, and duckduckgoing i went. P3 stands for "Power Disable Feature" it turns out, and lo and behold, all those thousands of returned drives were never defective at all.

SATA PWDIS has been around for a while; it's not just these HGST drives but also a number of WD shucks. You can buy cheap SATA power splitters (crimped or IDC, not molded) and rip out the 3.3v line (closest to the "L" in the connectors). Or you can use a Molex-to-SATA power splitter. Molex-powered backplanes also are immune to this issue. Kapton tape on the drives also works but is per-drive. If none of your power connectors have the 3.3v line, then you never have to worry about checking which drives use PWDIS.