[gentoo-user] USB NVMe connection problem

[gentoo-user] USB NVMe connection problem

[Peter Humphrey]

Greetings,

The journey continues...

I recently bought a Ugreen USB-3 external NVMe housing and a Samsung 990 Pro
4TB SSD. They worked fine for a few weeks, including yesterday when I used it
to back up my LAN server. Today, every attempt to connect it returns a dmesg
error: "Read Capacity(10) failed" and of course it can't be connected. I've
tried it on three machines, all with the same result.

My quandary is: how can I tell whether the Ugreen controller is not querying
the SSD, or the SSD is not answering it? Which component do I return as
faulty?

The full dmesg entry:

usb 3-7: new high-speed USB device number 7 using xhci_hcd
usb 3-7: New USB device found, idVendor=0bda, idProduct=9210, bcdDevice=20.01
usb 3-7: New USB device strings: Mfr=1, Product=2, SerialNumber=3
usb 3-7: Product: Ugreen Storage Device
usb 3-7: Manufacturer: Ugreen
usb 3-7: SerialNumber: 012938058E61
usb-storage 3-7:1.0: USB Mass Storage device detected
scsi host2: usb-storage 3-7:1.0
scsi 2:0:0:0: Direct-Access     Realtek  RTL9210          1.00 PQ: 0 ANSI: 6
sd 2:0:0:0: [sda] Read Capacity(10) failed: Result: hostbyte=DID_OK driverbyte=DRIVER_OK
sd 2:0:0:0: [sda] Sense Key : Illegal Request [current]
sd 2:0:0:0: [sda] Add. Sense: Invalid field in cdb
sd 2:0:0:0: [sda] 0 512-byte logical blocks: (0 B/0 B)
sd 2:0:0:0: [sda] 0-byte physical blocks
sd 2:0:0:0: [sda] Write Protect is off
sd 2:0:0:0: [sda] Mode Sense: 37 00 00 08
sd 2:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
sd 2:0:0:0: [sda] Attached SCSI disk

-- 
Regards,
Peter.

Re: [gentoo-user] USB NVMe connection problem

[Thomas Schweikle]

[-- Attachment #1: Type: text/plain, Size: 2281 bytes --]

Any second USB NVMe disk housing available? If yes, test the disk out in
there. If it works: you found out the disk housing is defective. If it does
not work: it may be the disk.
Do you have a second NVMe disk available? Switch disks. Then check if the
second disk works with this maybe defective disk housing. If it does, the
disk may be defective. If it does not, the housing is defective.
It may even be both are defective: the disk and the housing.

On Sun, Mar 30, 2025 at 2:45 PM Peter Humphrey <peter@prh.myzen.co.uk>
wrote:

> Greetings,
>
> The journey continues...
>
> I recently bought a Ugreen USB-3 external NVMe housing and a Samsung 990
> Pro
> 4TB SSD. They worked fine for a few weeks, including yesterday when I used
> it
> to back up my LAN server. Today, every attempt to connect it returns a
> dmesg
> error: "Read Capacity(10) failed" and of course it can't be connected. I've
> tried it on three machines, all with the same result.
>
> My quandary is: how can I tell whether the Ugreen controller is not
> querying
> the SSD, or the SSD is not answering it? Which component do I return as
> faulty?
>
> The full dmesg entry:
>
> usb 3-7: new high-speed USB device number 7 using xhci_hcd
> usb 3-7: New USB device found, idVendor=0bda, idProduct=9210,
> bcdDevice=20.01
> usb 3-7: New USB device strings: Mfr=1, Product=2, SerialNumber=3
> usb 3-7: Product: Ugreen Storage Device
> usb 3-7: Manufacturer: Ugreen
> usb 3-7: SerialNumber: 012938058E61
> usb-storage 3-7:1.0: USB Mass Storage device detected
> scsi host2: usb-storage 3-7:1.0
> scsi 2:0:0:0: Direct-Access     Realtek  RTL9210          1.00 PQ: 0 ANSI:
> 6
> sd 2:0:0:0: [sda] Read Capacity(10) failed: Result: hostbyte=DID_OK
> driverbyte=DRIVER_OK
> sd 2:0:0:0: [sda] Sense Key : Illegal Request [current]
> sd 2:0:0:0: [sda] Add. Sense: Invalid field in cdb
> sd 2:0:0:0: [sda] 0 512-byte logical blocks: (0 B/0 B)
> sd 2:0:0:0: [sda] 0-byte physical blocks
> sd 2:0:0:0: [sda] Write Protect is off
> sd 2:0:0:0: [sda] Mode Sense: 37 00 00 08
> sd 2:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't
> support DPO or FUA
> sd 2:0:0:0: [sda] Attached SCSI disk
>
> --
> Regards,
> Peter.
>
>
>
>
>

-- 
Thomas

[-- Attachment #2: Type: text/html, Size: 2842 bytes --]

Re: [gentoo-user] USB NVMe connection problem

[Michael]

[-- Attachment #1: Type: text/plain, Size: 2718 bytes --]

On Sunday, 30 March 2025 13:45:09 British Summer Time Peter Humphrey wrote:
> Greetings,
> 
> The journey continues...
> 
> I recently bought a Ugreen USB-3 external NVMe housing and a Samsung 990 Pro
> 4TB SSD. They worked fine for a few weeks, including yesterday when I used
> it to back up my LAN server. Today, every attempt to connect it returns a
> dmesg error: "Read Capacity(10) failed" and of course it can't be
> connected. I've tried it on three machines, all with the same result.
> 
> My quandary is: how can I tell whether the Ugreen controller is not querying
> the SSD, or the SSD is not answering it? Which component do I return as
> faulty?
> 
> The full dmesg entry:
> 
> usb 3-7: new high-speed USB device number 7 using xhci_hcd
> usb 3-7: New USB device found, idVendor=0bda, idProduct=9210,
> bcdDevice=20.01 usb 3-7: New USB device strings: Mfr=1, Product=2,
> SerialNumber=3
> usb 3-7: Product: Ugreen Storage Device
> usb 3-7: Manufacturer: Ugreen
> usb 3-7: SerialNumber: 012938058E61
> usb-storage 3-7:1.0: USB Mass Storage device detected

The above shows the USB device responded and consequently was detected by the 
kernel.


> scsi host2: usb-storage 3-7:1.0
> scsi 2:0:0:0: Direct-Access     Realtek  RTL9210          1.00 PQ: 0 ANSI: 6

and the controller handles SCSI commands, as expected.  So far, so good with 
respect to the USB controller.


> sd 2:0:0:0: [sda] Read Capacity(10) failed: Result: hostbyte=DID_OK
> driverbyte=DRIVER_OK sd 2:0:0:0: [sda] Sense Key : Illegal Request
> [current]

This "Sense Key" is the response from the block device.  It means some command 
data sent to it was deemed to be illegal and the device bailed out from 
executing the command.  More details to make sense of this message may be in 
the SCSI specification.


> sd 2:0:0:0: [sda] Add. Sense: Invalid field in cdb
> sd 2:0:0:0: [sda] 0 512-byte logical blocks: (0 B/0 B)
> sd 2:0:0:0: [sda] 0-byte physical blocks
> sd 2:0:0:0: [sda] Write Protect is off
> sd 2:0:0:0: [sda] Mode Sense: 37 00 00 08
> sd 2:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't
> support DPO or FUA sd 2:0:0:0: [sda] Attached SCSI disk

Does lsblk -a show the partition(s) on the device?

Does your Ugreen enclosure come with an external power supply and if yes, have 
you powered it up from the mains?  I'm thinking aloud here, this might be a 
matter of providing adequate power to it - but it does not explain why it 
suddenly started happening today.

You've already tried it on different PCs, so this problem is not isolated to a 
single PC.

Have you removed the SSD and connected it directly to the MoBo without any USB 
intermediary?  This will separate a USB Vs SSD hardware problem.

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [gentoo-user] USB NVMe connection problem

[Peter Humphrey]

On Sunday, 30 March 2025 14:19:47 British Summer Time Michael wrote:
> On Sunday, 30 March 2025 13:45:09 British Summer Time Peter Humphrey wrote:

--->8

> > sd 2:0:0:0: [sda] Read Capacity(10) failed: Result: hostbyte=DID_OK
> > driverbyte=DRIVER_OK sd 2:0:0:0: [sda] Sense Key : Illegal Request
> > [current]
> 
> This "Sense Key" is the response from the block device.  It means some
> command data sent to it was deemed to be illegal and the device bailed out
> from executing the command.  More details to make sense of this message may
> be in the SCSI specification.

Interesting.

> > sd 2:0:0:0: [sda] Add. Sense: Invalid field in cdb
> > sd 2:0:0:0: [sda] 0 512-byte logical blocks: (0 B/0 B)
> > sd 2:0:0:0: [sda] 0-byte physical blocks
> > sd 2:0:0:0: [sda] Write Protect is off
> > sd 2:0:0:0: [sda] Mode Sense: 37 00 00 08
> > sd 2:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't
> > support DPO or FUA sd 2:0:0:0: [sda] Attached SCSI disk
> 
> Does lsblk -a show the partition(s) on the device?

It shows a disk of size 0. (I keep forgetting to try lsblk, which I haven't 
had to use until this exercise began.)

> Does your Ugreen enclosure come with an external power supply

Nope.

> You've already tried it on different PCs, so this problem is not isolated to
> a single PC.
> 
> Have you removed the SSD and connected it directly to the MoBo without any
> USB intermediary?  This will separate a USB Vs SSD hardware problem.

I've a nasty feeling I damaged the SSD by applying too much force while 
installing it. Cracked track, or something. If so, I don't feel I should 
return it as faulty.

[Slumps shoulders...]

Can I justify starting again and spending yet more hundreds? That's the 
question. Especially as the spinning disk it was to replace hasn't failed yet.

-- 
Regards,
Peter.

Re: [gentoo-user] USB NVMe connection problem

[Michael]

[-- Attachment #1: Type: text/plain, Size: 717 bytes --]

On Sunday, 30 March 2025 16:05:29 British Summer Time Peter Humphrey wrote:

> I've a nasty feeling I damaged the SSD by applying too much force while
> installing it. Cracked track, or something. If so, I don't feel I should
> return it as faulty.
> 
> [Slumps shoulders...]
> 
> Can I justify starting again and spending yet more hundreds? That's the
> question. Especially as the spinning disk it was to replace hasn't failed
> yet.

Regretfully accidents do happen, as I well know first hand.  ;-)

As I suggested you can connect briefly the SSD directly on a MoBo port to 
confirm if there is a problem with it.  If it shows up and the partitions are 
all in place, then the fault must be with the USB enclosure.

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[OT} Re: [gentoo-user] USB NVMe connection problem

[Peter Humphrey]

On Sunday, 30 March 2025 16:05:29 British Summer Time I wrote:

> Can I justify starting again and spending yet more hundreds?

I've taken the plunge and ordered new SSD and enclosure.

I can't justify the cost of course, but this has been my life, ever since my 
first hardware maintenance course in 1972 (Honeywell DAP16 for temperature 
analysis at a new AGR nuclear power station in Scotland), through all the 
years until I led the team who maintained the software in 10 languages on the 
15-mainframe Cyber 960 system that runs the National Grid. Well, it did then, 
but that was in the '90s. Since then of course the landscape has changed 
utterly, what with the lunatic mania for 'sustainable' energy sources.

I retired in 1989 but kept my interest in all things computer. I'm sure it 
will be with me until my last day. 

Anyway, I just wanted to show how much I couldn't resist buying the new kit.

-- 
Regards,
Peter.