Seagate 4TB SAS ST4000NM0025 EXOS 7E8 Firmware? Drives possibly removed from Hitachi VSP?

[sb1racing]

Does anyone have firmware available for the Seagate ST4000NM0025 drives?

I bought some that are using some crazy firmware and where 520. I reformatted them to 512 with sg_format without any issues but now can't seem to write anything to them. In the past a firmware update usually resolves this but I can't seem to find one for these drives?!?

parted:
(parted) mklabel gpt
Error: Input/output error during write on /dev/sda

fdisk:
The partition table has been altered.
Calling ioctl() to re-read partition table.
/dev/sda: fsync device failed: Input/output error

sg_readcap:
Read Capacity results:
Protection: prot_en=0, p_type=0, p_i_exponent=0
Logical block provisioning: lbpme=0, lbprz=0
Last logical block address=7814037167 (0x1d1c0beaf), Number of logical blocks=7814037168
Logical block length=512 bytes
Logical blocks per physical block exponent=0
Lowest aligned logical block address=0
Hence:
Device size: 4000787030016 bytes, 3815447.8 MiB, 4000.79 GB

smartctl:
=== START OF INFORMATION SECTION ===
Vendor: SEAGATE
Product: DKS2G-H4R0SS
Revision: 6F08
Compliance: SPC-3
User Capacity: 4,000,787,030,016 bytes [4.00 TB]
Logical block size: 512 bytes
LU is fully provisioned
Rotation Rate: 7200 rpm
Form Factor: 3.5 inches
Logical Unit id: 0x5000c500a6815b07
Serial number: ZC173DGD0000R851KNWH
Device type: disk
Transport protocol: SAS (SPL-3)
Local Time is: Fri Oct 29 13:36:39 2021 EDT
SMART support is: Available - device has SMART capability.
SMART support is: Enabled
Temperature Warning: Disabled or Not Supported

Any help would be greatly appreciated!!!
Thanks!!

 

[itronin]

so I don't see any protection types enabled on the drive - so it is not that.
The symptoms you are describing are pretty much those of a drive that was not properly exported from a Hitachi VSP system for use in non VSP deployments.

Your drive type does not show up as what I'd call "typical" for a VSP - those are usually 2.5" 300GB-1.92TB 10k-15k drives, both seagate and HGST labeled. However your product ID smells like VSP based on what I've personaslly tried to get working (0SS was always there in the id). A quick google search says your drive type does look like it may be from an HGST HUS system - I dont' know anything about those but would not be surprised if it also implemented the "funk" that makes them not work right in simple HBA deployments.

the VSP drives have "funk" in their firmware and implement some sort of delayed or controller triggered and managed write scheme. If these drives came out of a VSP you may be stuck and your best option is to return for refund and buy something. There has been very limited success getting new firmware into VSP drives that made them operational. There was a member who reportedly succeeded. they required folks to send drives to them (in Russia I think) where they did some low level block surgery to rewrite the base drive firmware to take "normal firmware".

You might take a look at this thread for more ideas. I would not be put off by "Hitachi" in the thread title, you go through and there are references to both Seagate (Cobra) drives that did not work and could not be updated as well as some that could. It may give you ideas etc.

something to consider: level of effort to "maybe" make them work versus (if this is still an option) returning them for a refund, spending a little bit more for drives from say an EMC, netapp or HP 3PAR that should just need to be reblocked, long format, and then ready to use.

 

[sb1racing]

Thanks for the info! Title has been updated. I did notice Hitachi was the only thing coming up when I used Google for "DKS2G-H4R0SS"... They have the Seagate Exos labels on them so I was confused by the search results. I can swap them out for different drives if I can't get these working. Just figured it was worth a shot before shipping them back

 

[enthus]

I stumbled on this thread with the exact same problem, exact same type of drive. Then I realized I have the same drive that you returned, they shipped it to me mid November. Serial number ZC173DGD0000R851KNWH. I have 4 of these from the same seller, cannot write to the drives at all but they pass Seagates bootable USB checks 100%.

Image of serial number.

It appears HP have a firmware for this type of drive, but it feels a bit risky to start randomly flashing.

HPE Support Center

support.hpe.com

It also appears that using a different serial number from another ST4000NM0025 1V4207 (Picture | SN: ZC1BFPDW) I was able to goad Seagates download tool into providing me with a firmware file on this page: Exos 7E | Seagate Support Canada
Enterprise Capacity 7200.5 SAS STD 512N Firmware Update
Enterprise Capacity 7200.5 SAS STD 512N (MakaraBP) firmware update

Perhaps this is the way to go?

Edit: So far all I get when I attempt to flash the firmware is "Firmware Download failed".

Edit: With some verbose logging I've narrowed the firmware update issue to this error:
ASC & ASCQ: 26h - 99h = Vendor specific ascq code
FRU: 3Fh = Vendor Specific

Which according to this document https://www.seagate.com/files/staticfiles/support/docs/manual/Interface manuals/100293068j.pdf means 26 99 Invalid Field Parameter—Firmware Tag.

If the firmware tag can be edited in a hex editor maybe it's doable. I've message the seller and I have 4 days to start the return process.

There is also hope in using an open source version of SeaChest to force the firmware. GitHub - Seagate/openSeaChest: Cross platform utilities useful for performing various operations on SATA, SAS, NVMe, and USB storage devices.

 

[nihonjin]

I stumbled on this thread with the exact same problem, exact same type of drive. Then I realized I have the same drive that you returned, they shipped it to me mid November. Serial number ZC173DGD0000R851KNWH. I have 4 of these from the same seller, cannot write to the drives at all but they pass Seagates bootable USB checks 100%.

Image of serial number.

It appears HP have a firmware for this type of drive, but it feels a bit risky to start randomly flashing.

HPE Support Center

support.hpe.com

It also appears that using a different serial number from another ST4000NM0025 1V4207 (Picture | SN: ZC1BFPDW) I was able to goad Seagates download tool into providing me with a firmware file on this page: Exos 7E | Seagate Support Canada
Enterprise Capacity 7200.5 SAS STD 512N Firmware Update
Enterprise Capacity 7200.5 SAS STD 512N (MakaraBP) firmware update

Perhaps this is the way to go?

Edit: So far all I get when I attempt to flash the firmware is "Firmware Download failed".

Edit: With some verbose logging I've narrowed the firmware update issue to this error:
ASC & ASCQ: 26h - 99h = Vendor specific ascq code
FRU: 3Fh = Vendor Specific

Which according to this document https://www.seagate.com/files/staticfiles/support/docs/manual/Interface manuals/100293068j.pdf means 26 99 Invalid Field Parameter—Firmware Tag.

If the firmware tag can be edited in a hex editor maybe it's doable. I've message the seller and I have 4 days to start the return process.

There is also hope in using an open source version of SeaChest to force the firmware. GitHub - Seagate/openSeaChest: Cross platform utilities useful for performing various operations on SATA, SAS, NVMe, and USB storage devices.

sorry for resurrecting this thread but, did you finally success flashing them?

 

[Tickety-boo]

sorry for resurrecting this thread but, did you finally success flashing them?

I have a kit of SCSITOOLS tools that it may be solve your problem.

 

[nihonjin]

I have a kit of SCSITOOLS tools that it may be solve your problem.

Thanks, is it tested?

 

[Tickety-boo]

No, I don't have this model yet. But I did use the tool's latest ST 10K V8 unlock feature.

 

[nihonjin]

I'll give it a try

Thanks.

 

[ownhost]

Hello
If the disc is still there, you can make the disc usable with the original firmware update. This is how I solved Hitachi Storage Disks.

 

[Adeel Akram]

Hello
If the disc is still there, you can make the disc usable with the original firmware update. This is how I solved Hitachi Storage Disks.

I have several of these disks.
Can you suggest how to install original firmware on them?
Which Hitachi Storage Disks you managed to make work in other systems?

 

[Adeel Akram]

I'll give it a try

Thanks.

Hi, Were you ever able to make the disks work?

 

[Adeel Akram]

I stumbled on this thread with the exact same problem, exact same type of drive. Then I realized I have the same drive that you returned, they shipped it to me mid November. Serial number ZC173DGD0000R851KNWH. I have 4 of these from the same seller, cannot write to the drives at all but they pass Seagates bootable USB checks 100%.

Image of serial number.

It appears HP have a firmware for this type of drive, but it feels a bit risky to start randomly flashing.

HPE Support Center

support.hpe.com

It also appears that using a different serial number from another ST4000NM0025 1V4207 (Picture | SN: ZC1BFPDW) I was able to goad Seagates download tool into providing me with a firmware file on this page: Exos 7E | Seagate Support Canada
Enterprise Capacity 7200.5 SAS STD 512N Firmware Update
Enterprise Capacity 7200.5 SAS STD 512N (MakaraBP) firmware update

Perhaps this is the way to go?

Edit: So far all I get when I attempt to flash the firmware is "Firmware Download failed".

Edit: With some verbose logging I've narrowed the firmware update issue to this error:
ASC & ASCQ: 26h - 99h = Vendor specific ascq code
FRU: 3Fh = Vendor Specific

Which according to this document https://www.seagate.com/files/staticfiles/support/docs/manual/Interface manuals/100293068j.pdf means 26 99 Invalid Field Parameter—Firmware Tag.

If the firmware tag can be edited in a hex editor maybe it's doable. I've message the seller and I have 4 days to start the return process.

There is also hope in using an open source version of SeaChest to force the firmware. GitHub - Seagate/openSeaChest: Cross platform utilities useful for performing various operations on SATA, SAS, NVMe, and USB storage devices.

Hi, were you ever able to make these disks work?

 

[Adeel Akram]

Does anyone have firmware available for the Seagate ST4000NM0025 drives?

I bought some that are using some crazy firmware and where 520. I reformatted them to 512 with sg_format without any issues but now can't seem to write anything to them. In the past a firmware update usually resolves this but I can't seem to find one for these drives?!?

parted:
(parted) mklabel gpt
Error: Input/output error during write on /dev/sda

fdisk:
The partition table has been altered.
Calling ioctl() to re-read partition table.
/dev/sda: fsync device failed: Input/output error

sg_readcap:
Read Capacity results:
Protection: prot_en=0, p_type=0, p_i_exponent=0
Logical block provisioning: lbpme=0, lbprz=0
Last logical block address=7814037167 (0x1d1c0beaf), Number of logical blocks=7814037168
Logical block length=512 bytes
Logical blocks per physical block exponent=0
Lowest aligned logical block address=0
Hence:
Device size: 4000787030016 bytes, 3815447.8 MiB, 4000.79 GB

smartctl:
=== START OF INFORMATION SECTION ===
Vendor: SEAGATE
Product: DKS2G-H4R0SS
Revision: 6F08
Compliance: SPC-3
User Capacity: 4,000,787,030,016 bytes [4.00 TB]
Logical block size: 512 bytes
LU is fully provisioned
Rotation Rate: 7200 rpm
Form Factor: 3.5 inches
Logical Unit id: 0x5000c500a6815b07
Serial number: ZC173DGD0000R851KNWH
Device type: disk
Transport protocol: SAS (SPL-3)
Local Time is: Fri Oct 29 13:36:39 2021 EDT
SMART support is: Available - device has SMART capability.
SMART support is: Enabled
Temperature Warning: Disabled or Not Supported

Any help would be greatly appreciated!!!
Thanks!!

Hi, were you ever able to make these disks work?

 

[enthus]

Hi, were you ever able to make these disks work?

Yes, I hired a sketchy russian to remote into a temporary computer I set up and he did some firmware magic to the drives. They have been running now for 3 years flawlessly and I am about to replace them with 12TB.

 

[nihonjin]

Yes, I hired a sketchy russian to remote into a temporary computer I set up and he did some firmware magic to the drives. They have been running now for 3 years flawlessly and I am about to replace them with 12TB.

No serial connection required? just by SAS interface? would love to know how...

 

[enthus]

No serial connection required? just by SAS interface? would love to know how...

Yes. But his requirements were a Remote Desktop connection on a Windows 7 32Bit machine. I complied and bashed the boot drive with a hammer when we were done. Not very comfortable to run such an old os and expose RDP to the internet. I also ran it on a different vlan just to be a little safer, but the whole thing was super dodgy. I was very happy it worked out though.

Because it was RDP I was unable to see what he did. I paid him $5 per drive if memory serves.

 

[jabuzzard]

Just posting because people are likely to stumble over this thread when they have similar problems with drives removed from a VSP system. There is a solution that works for ALL spinning drives taken from a VSP, that involves setting a magic bit in a undocumented mode page of the drive. Full writeup on the VSP situation along with a Bash script to automate the process of formatting and setting the magic bit on drives. Really handy if you have a lot of drives to process.

GitHub - digitalcabbage/vspenable: Enable SCSI_WRITE on hard drives removed from a HGST VSP storage system

Enable SCSI_WRITE on hard drives removed from a HGST VSP storage system - digitalcabbage/vspenable

github.com

 

[jabuzzard]

NetApp OEM Seagate drives like that DKS2G-H4R0SS are notoriously stubborn when shifting from 520 to 512 bytes because of the embedded T10-PI metadata blocks. Changing the block size alone with `sg_format` leaves the protection interval active, which triggers the `Input/output error` on fsync you're seeing in fdisk. You'll need a full low-level format command specifying zero protection type to make it writable under standard Linux kernels.

Drives from a VSP are a different, stubborn though. No amount of sg_format will make them work, period. You either need to flash different firmware or fiddle with the magic bit on mode page 0x38. Older Seagate drives, specifically ones with a SAS-1 or SAS-2 interface removed from a VSP, can be flashed with generic firmware. On more modern drives, aka ones with a SAS-3 or better interface, they are vendor-locked and cannot be flashed with generic firmware. The ST4000NM0025 of this thread is a SAS-3 12Gbps drive and cannot be easily flashed with generic firmware. However, a simple reformat to a 512-byte block size followed by unlocking the SCSI_WRITE command in mode page 0x38 will get them working.

Note that the magic bit in mode page 0x38 is a relatively new discovery. It was first hinted at in a single thread on the HDD Guru forums back in 2023, but it was incomplete: it worked only for Seagate drives and was a manual, error-prone process. The script linked above automates the process and works on both Seagate and HGST/Western Digital drives, tested on multiple different drive models (around 10) from both Seagate and HGST.