HGST OEM LOCKED SAS
- Thread starter arsalkhan
- Start date
Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
Your symptoms, the firmware version and the model number make me think these are drives pulled from a Hitachi VSP without being off-boarded. If they are then firmware needs to be replaced. there is an article or two on the forum about various peoples' adventures trying to get them working. the only success I am aware of is a person in Russia who can manually rework the firmware.
If you know someone with an operable VSP then you could try adding them and use the Hitachi tools to off-board them.
If you know someone with an operable VSP then you could try adding them and use the Hitachi tools to off-board them.
I'm sorry I never used the service. Here's one of the STH threads on this topic with a link to an hdguru forum thread.
You need special firmware software applications that can fix the firmware config. These firmware applications are not generally available.
and what's the magic bit and what's the undocumented mode page?
and is this "any drive" or only specific to the drives you are testing against? what models? still ton's of those VSP not-properly-exported drives floating around - especially 1 and 2 tb 2.5" spinners
and is this "any drive" or only specific to the drives you are testing against? what models? still ton's of those VSP not-properly-exported drives floating around - especially 1 and 2 tb 2.5" spinners
Doing it by hand is incredibly fiddly and given the potential to brick drives I think it's best everyone waits till I have a robust script to do the job automatically. Besides you wouldn't want to do a couple hundred drives by hand it would take days. I only discovered the complete solution yesterday. The good news is there is definitely a way to get these drives working quickly and at no cost, though donations welcome 
It will, I think, work with any drive though I cannot be sure as I have only tested a subset of drives. Though that is quite a few and it has worked with all the models I have tried. Note I have only ever seen Seagate and HGST drives from a VSP. The following drive models are known to work though the first three Seagate models are easy enough to flash with generic Seagate firmware to get around the SCSI_WRITE_AND_VERIFY issue, which is probably a better idea anyway. I have included in brackets valid serial numbers (gleaned from eBay pics) to download the generic firmware from the Seagate website.
ST9900805SS (6XS2BPEC)
ST900MM0006 (S0N1WQH7)
ST4000NM0023 (Z1Z8CLYA)
ST900MM0168
ST1200MM0009
ST4000NM0025
ST6000NM0034
ST10000NM002G
ST10000NM013G
HUC109090CSS600
HUH721010AL5204
Note the ST1200MM0009 drives I have access to had been made to "appear" as 900GB drives presumably due to them being replacements for HUC109090CSS600 drives. On formatting down to 512b block size they have reverted to 1.2TB in size, and have been successfully unlocked and now behave as 1.2TB 10k RPM drives. An interesting point is the undocumented code page disappears on the drives I have been able to flash to generic firmware. It should also be noted that the mode page is different on every single drive that I have manually looked at. You have to grab the current mode page, flip the relevant bit, convert it to a hexadecimal string and then use sg_wr_mode to write it back out again.
Patient is a virtue that is pleasant to behold. She sits upon a monument and does as she is told.
It will, I think, work with any drive though I cannot be sure as I have only tested a subset of drives. Though that is quite a few and it has worked with all the models I have tried. Note I have only ever seen Seagate and HGST drives from a VSP. The following drive models are known to work though the first three Seagate models are easy enough to flash with generic Seagate firmware to get around the SCSI_WRITE_AND_VERIFY issue, which is probably a better idea anyway. I have included in brackets valid serial numbers (gleaned from eBay pics) to download the generic firmware from the Seagate website.
ST9900805SS (6XS2BPEC)
ST900MM0006 (S0N1WQH7)
ST4000NM0023 (Z1Z8CLYA)
ST900MM0168
ST1200MM0009
ST4000NM0025
ST6000NM0034
ST10000NM002G
ST10000NM013G
HUC109090CSS600
HUH721010AL5204
Note the ST1200MM0009 drives I have access to had been made to "appear" as 900GB drives presumably due to them being replacements for HUC109090CSS600 drives. On formatting down to 512b block size they have reverted to 1.2TB in size, and have been successfully unlocked and now behave as 1.2TB 10k RPM drives. An interesting point is the undocumented code page disappears on the drives I have been able to flash to generic firmware. It should also be noted that the mode page is different on every single drive that I have manually looked at. You have to grab the current mode page, flip the relevant bit, convert it to a hexadecimal string and then use sg_wr_mode to write it back out again.
Patient is a virtue that is pleasant to behold. She sits upon a monument and does as she is told.
Last edited:
What real world drive is the DKR5E-J1R2SS? My guess is that it is a HGST 2.5" drive but what model?
After some Googling, they are HUC101212CSS600 drives, so HGST 1.2TB, 10k RPM drives.
So the format of the model name of the drive as returned by sginfo for the VSP drives does have some structure to it DK[SR]\w\w-[HJ]\w\w\wSS
S = Seagate
R = HGST
H = 3.5"
J = 2.5"
where \w is uppercase only. I suspect the last three digits represent the capacity. If it is three digits, then it is the capacity in GB, so a 900 is a 900GB drive. If there is an R, that is a decimal point in TB capacity, so 1R2 is 1.2TB, 4R0 is 4 TB, 6R0 is 6 TB, and 10R is 10 TB.
Last edited:
That is not accurate the letter after the dash refers to the rotational speed of the drive H=7200, J=10k and K=15k. The physical size of the drive comes from the fourth character which is 2=3.5" and 5=2.5"
The fifth letter appears to refer to the generation of the drive and is increasing over time. That is new generations of drive have a letter than it higher in the alphabet than old ones. Presumably you need a lookup table to know exactly what they refer to.
Just some more random updates on the model decoding. The last two letters are the interface type
SS = SAS
AT = SATA
NC = NVMe
I suspect if you went far back in time enough there would be a code for FIbre Channel drives. Additional manufacturer codes
B = Toshiba/Kioxia
M = Western Digital
Speed
H = 7200 rpm
J = 10k rpm
K = 15k rpm
M = SSD
I suspect that Read intensive, mixed use, write intensive might have different letters, because the first two letters indicate the drive type
DK = spinning disk
SF = flash
SS = SAS
AT = SATA
NC = NVMe
I suspect if you went far back in time enough there would be a code for FIbre Channel drives. Additional manufacturer codes
B = Toshiba/Kioxia
M = Western Digital
Speed
H = 7200 rpm
J = 10k rpm
K = 15k rpm
M = SSD
I suspect that Read intensive, mixed use, write intensive might have different letters, because the first two letters indicate the drive type
DK = spinning disk
SF = flash
The script and some description of everything can be found here
github.com
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
great detective work and a very nice write and its wonderful to see someone dropping in sources and history. great solution. I think I just found a box with a few VSP drives (cobra with the joker code) left over from my experiments back when about 8 years ago now.
If I really did I hope to give your method a go!
If I really did I hope to give your method a go!
Thanks. It is the only place on the internet that pulls all the knowledge about reusing VSP drives into a single place. I have added the script that I used for flashing the SAS-2/6Gbps Seagate drives and made a few small tidy ups in the code. The flashing is for me of historical interest because all the drives I could flash have been flashed bar a handful I have kept for experimentation I have been fiddling with overcoming the vendor locks with the serial port on the Seagate drives. So far I only have a small pile of bricked drives, but if I can make it work and flash Dell firmware onto the drives they would be useful in MD storage arrays at work.
I have been fiddling with overcoming the vendor locks with the serial port on the Seagate drives. So far I only have a small pile of bricked drives, but if I can make it work and flash Dell firmware onto the drives they would be useful in MD storage arrays at work.