The quest for the HGST UltraStar SN260 firmware updates...

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

joerambo

New Member
Aug 30, 2023
28
9
3
nvme create-ns /dev/nvme0 --nsze=15016584882 --ncap=15016584882 --flbas=0 --dps=0 --nmic=0
That's what i mean by "vendor" magic, but i'd start with basics, like what LLM recommend:

nvme create-ns /dev/nvme0 -s $(nvme id-ctrl /dev/nvme0 | grep tnvmcap | awk '{print $3}') -b 4096
 

zxc_ee

Member
Apr 10, 2024
48
37
18
49
.be / .nl / .ee
That's what i mean by "vendor" magic, but i'd start with basics, like what LLM recommend:

nvme create-ns /dev/nvme0 -s $(nvme id-ctrl /dev/nvme0 | grep tnvmcap | awk '{print $3}') -b 4096
root@x:~# nvme create-ns /dev/nvme0 -s $(nvme id-ctrl /dev/nvme0 | grep tnvmcap | awk '{print $3}') -b 4096
NVMe status: Unknown(0x47c5)

any value will give same result :(
 

zxc_ee

Member
Apr 10, 2024
48
37
18
49
.be / .nl / .ee
And nvme list-subsys ?
Code:
root@x:~# nvme list-subsys
nvme-subsys0 - NQN=nqn.2017-03.com.wdc:nvme-solid-state-drive. VID:1C58.        MN:HUSMR7676BDP3Y1     .SN:SDM000088F73
\
 +- nvme0 pcie 0000:08:00.0 live
root@x:~#
root@x:~#
root@x:~# nvme list-subsys -vv
nvme-subsys0 - NQN=nqn.2017-03.com.wdc:nvme-solid-state-drive. VID:1C58.        MN:HUSMR7676BDP3Y1     .SN:SDM000088F73
\
 +- nvme0 pcie 0000:08:00.0 live
 

ElCoyote_

Active Member
Jul 22, 2016
248
190
43
also nvme id-ctrl reporting 1, is it ok , just one ns??

Code:
nvme id-ctrl /dev/nvme0 | grep ^nn
nn        : 1
[/QUOTE]

is it possible that your drive is in 'diagnostics' mode? those cards usually have 4 namespaces.
From the output you posted, this seems like a Cisco UCS drive: UCSC-NVMEHW-H76800239, make sure to load the Cisco firmware for it.
 

zxc_ee

Member
Apr 10, 2024
48
37
18
49
.be / .nl / .ee
also nvme id-ctrl reporting 1, is it ok , just one ns??

Code:
nvme id-ctrl /dev/nvme0 | grep ^nn
nn        : 1
is it possible that your drive is in 'diagnostics' mode? those cards usually have 4 namespaces.
From the output you posted, this seems like a Cisco UCS drive: UCSC-NVMEHW-H76800239, make sure to load the Cisco firmware for it.
[/QUOTE]


yes, it's cisco ucs drive, i tried to load padded file, but without much luck.
also tried different motherboards/server,incl. supermicro. syslog is full of messages:

Code:
nvme nvme0: 4/0/0 default/read/poll queues
nvme nvme0: resetting controller due to AER
 

zxc_ee

Member
Apr 10, 2024
48
37
18
49
.be / .nl / .ee
case closed, managed to get device out of diag mode with windows dm tool, ver 2.12.1 , win 2016.
now device visible and functional in windows and linux
 

vitamins

New Member
Oct 27, 2023
8
13
3
I was able to rescue ~30 dell branded SN200's from diagnostic mode using dm-cli and the ' --clear-diag-data' command from above in windows. I left the dell firmware on all except one, which I flashed generic firmware on (KNGND122). I ran a read/write test in HDSentinel over a weekend on two drives, read and wrote about 65TB. The generic firmware drive went down in health by 1%. Not sure if that was a firmware issue or if the drive was going to go down regardless. I'll also say that the above method set the byte sector to 4096 rather than dells 520 b/s, also resets most smart info from what I can tell. TBW, power on hours, start stop count will all be reset to 0. Health remained the same though at ~95%.
 

pimposh

hardware pimp
Nov 19, 2022
434
270
63
If it is not happening already, now we can expect „brand new” drives on ebay. Nice, but not.
 
Last edited:

luckylinux

Well-Known Member
Mar 18, 2012
1,613
518
113
I have ordered the HGST Ultrastar SN100 SSD HUSPR3216ADP301. Can Someone help find firmware updates?

EDIT:



From above download that firmware archive "HGST-UltraStar-SN150-HHHL.zip" seem to be also compatible with SN100 series drives. The zip file is titled SN150 though, I hope it will not brick my drive.
Did you manage to get it working on the SN100?

I recently ordered a HGST 3.2TB HUSPR3232ADP301(SN100) and I couldn't find any public links available for Firmware Downloads (except the Links here and your Post).
 

serverfan

New Member
Jan 26, 2022
7
0
1
Hopping in. I'm the one who bought 20 of them and trying to pair with Dell R640 (nvme cables are installed).
All of them are falling in diagnostics mode on each boot. I was able to cancel diagnostics mode and upgrade a firmware to KNGND122 using RHEL8.
However, they failing to diag mode and vanishes from iDRAC physical disks section.
Fun fact: hdm util in Ubuntu can't see these drives with "scan" command. While booting dmesg shows

dmesg | grep nvme
[ 9.570101] nvme nvme0: pci function 0000:86:00.0
[ 9.572092] nvme nvme1: pci function 0000:87:00.0
[ 9.578342] nvme nvme2: pci function 0000:88:00.0
[ 9.581914] nvme nvme3: pci function 0000:89:00.0
[ 12.163999] nvme nvme0: 112/0/0 default/read/poll queues
[ 12.171845] nvme nvme2: 112/0/0 default/read/poll queues
[ 12.188559] nvme nvme3: 112/0/0 default/read/poll queues
[ 12.188897] nvme nvme1: 112/0/0 default/read/poll queues|
... then repeatedly error messages about AER appear


UPD: Drives are Dell EMC branded: EMC 005052504 / HUSMR7638BDP3Y1 WD HGST Ultrastar SN200
Before firmwaring to KNGND122.bin they had KNECD116 version/
Any thoughts?
 
Last edited:

ElCoyote_

Active Member
Jul 22, 2016
248
190
43
@serverfan for these drives in 14th gen Dells, they need to be on firmware 122 (it's the only one which works if your BIOS is later than 2.13.z).
they might go into diags mode and in which case the only way I found the recover them is to dump the diags:
Code:
# dm-cli capture-diagnostics  --file /tmp/p0 --clear-diag-data -p /dev/nvme0
# dm-cli capture-diagnostics  --file /tmp/p1 --clear-diag-data -p /dev/nvme1
# dm-cli capture-diagnostics  --file /tmp/p2 --clear-diag-data -p /dev/nvme2
# dm-cli capture-diagnostics  --file /tmp/p3 --clear-diag-data -p /dev/nvme3
 

serverfan

New Member
Jan 26, 2022
7
0
1
@serverfan for these drives in 14th gen Dells, they need to be on firmware 122 (it's the only one which works if your BIOS is later than 2.13.z).
they might go into diags mode and in which case the only way I found the recover them is to dump the diags:
Yep, I found bunch of topics related to SN200 where you posted a solution to recover them, thank you!

So i came up to recovering from diag mode, shutdown server and then booting as usual into Ubuntu, it works so far... However, probably will look for opportunity to switch from them. Probably to sell in a summer I guess.
 

LodeRunner

Active Member
Apr 27, 2019
568
244
43
Of the threads about the SN200's on the internet, this one has the best info I've found. I got the drives out of diag mode using HDM in Windows, but can't update the firmware. The drives appear to be HGST, no other branding to indicate OEM. I have 2.5" 3.84 TB SN200's, model HUSMR7638BDP3Y1, which was in the text file of the ZIP files you posed on Dropbox. Server they're installed in is a Dell R640 with latest BIOS and iDRAC firmware (updated before acquisition of the drives).

Device info:
Code:
[@nvme0]
  Product Name                  = Ultrastar SN200
  Device Type                   = NVMe Controller
  Device Path                   = \\?\pci#ven_1c58&dev_0023&subsys_00231c58&rev_02#000cca0c015c7a8000#{2accfe60-c130-11d2-b082-00a0c91efb8b}
  UID                           = 1C58SDM0000484D3HUSMR7638BDP3Y10023
  Alias                         = @nvme0
  Vendor Name                   = HGST
  Model Name                    = HUSMR7638BDP3Y1
  Vendor ID                     = 1C58
  Device ID                     = 0023
  Firmware Version              = KNECD112
  Default Capacity              = 3840000000000
  Capacity                      = 3840755982336
  NVM Subsystem UID             = 1C58SDM0000484D3HUSMR7638BDP3Y1
  Namespace Count               = 1
  Multipath Support             = false
  Multipath Mode                = Single Port
  Namespace Sharing Support     = false
  Encryption Support            = false
  Encryption Mode               = No Encryption
  UEFI Driver Version           = 1c58-0023-105c
  SBL Version                   = KNEGZ107_SFF2_R01_Q16_S01_N040_D3_G04_X4
  PCI Domain Number             = 0000
  PCI Bus Number                = 136
  PCI Device Number             = 0
  PCI Function Number           = 0
  PCI Max Link Speed            = PCIe 3.0 (8.0 GT/s)
  PCI Max Link Width            = x4
  PCI Current Link Speed        = PCIe 3.0 (8.0 GT/s)
  PCI Current Link Width        = x4

Firmware info:
Code:
[@nvme0]
  Product Name             = Ultrastar SN200
  Device Type              = NVMe Controller
  Device Path              = \\?\pci#ven_1c58&dev_0023&subsys_00231c58&rev_02#000cca0c015c7a8000#{2accfe60-c130-11d2-b082-00a0c91efb8b}
  UID                      = 1C58SDM0000484D3HUSMR7638BDP3Y10023
  Alias                    = @nvme0
  Running Firmware Version = KNECD112 (Loaded from Slot 2)
  Slot 1 (Read-only)       = KNEGZ107
  Slot 2                   = KNECD112 (Next Running Firmware Slot)
  Slot 3                   = KNEGZ107
  Slot 4                   = KNEGZ107
  Slot 5                   = KNECD10B
Attempting to update the firmware results in this:
Code:
PS C:\Users\Administrator\Downloads> hdm manage-firmware --load -f KNGND122.bin --slot 5 -a "@nvme0"
[Errors]
  Device firmware version is not compatible with this operation
    Device Reference = @nvme0

Results for manage-firmware: Operation failed.
Same if I try using KNGND100.bin and KNGND110.bin.

Drive state:
Code:
[@nvme0]
  Product Name                = Ultrastar SN200
  Device Type                 = NVMe Controller
  Device Path                 = \\?\pci#ven_1c58&dev_0023&subsys_00231c58&rev_02#000cca0c015c7a8000#{2accfe60-c130-11d2-b082-00a0c91efb8b}
  UID                         = 1C58SDM0000484D3HUSMR7638BDP3Y10023
  Alias                       = @nvme0
  Device Status               = Ready
  Thermal Throttling Status   = Disabled
  Beacon LED Enabled          = false
  Life Gauge                  = 97
  System Area Percentage Used = 2
  Power Consumption           = 8.439000

Results for get-state: Operation succeeded.
The device properly shows in Device Manager as WD Ultrastar SN2xx PCIe SSD Controller, and as a disk labeled NVMe HUSMR7638BDP3Y1. Disk Management sees them and I can initialize them, so they appear to be working, other than blocking firmware updates. If I boot to TrueNAS, they immediately switch to diag mode.

Any tips on the path forward? Do I need other intermediate fw versions?