Have scammers finally circumvented ValiDrive's tests ?

[astrostar]

I recently purchased two 4Tb SSD drives from Aliexpress and was pleased to see that ~ValiDrive reported both drives as being 4Tb - great start, or so I thought !
I have around 650Gb stored on an internal SAS drive which I wanted to transfer onto one of the SSD drives so started copying the files using TeracCopy. I only have USB 2 ports so the transfers were running at around 28Mb/s. I left the computer running but when I checked around an hour or so later TeraCopy had stopped copying with around 300Gb transferred and was showing a load of problems with the verification process.
I then tried the second SSD drive with the exact same problem so I'm guessing the SSD drives are not 4Tb but probably nearer 300Gb and if this is the case then it would appear that the scammers have found a way to fool the ValiDrive software.
Anyone have any thoughts on this ?

 

[DanR]

@astrostar AliExpress appears to be a Chinese company. So . . . I would not be surprised. Steve is planning on ValiDrive v2 eventually.

Have you tried rerunning ValiDrive v1 again now that you have copied data to the drives?

 

[peterblaise]

Free ValiDrive can identify some bad drives, but does not
exhaustively identify any drive as good - a conundrum, I
know.

So, also run free h2testw https://h2testw.org/ - it tests
formatted volumes - format each drive as one blank
exFAT or NTFS volume, then test.

Tell us what you find, then return 'em if they are really
bad.

 

[ColbyBouma]

How much did you pay for the drives? Are you willing to provide a link to the AliExpress page for them?

 

[peterblaise]

Though NOT designed to test intentionally lying drives,
SpinRite 6.1 Level 5 may also reveal the transition
between 'good' and fake sectors, either by confirming
errors, or by going artificially speedy across fake areas,
but access via USB is potentially also compromised by
SpinRite 6.1 depending on BIOS and DOS - ValiDrive
takes advantage of Windows USB drivers.

An alternative may be to 'look' at the drives via graphing
software, such as free HDDScan https://hddscan.com/
which may reveal an artificially speedy ( fake ) response
on the, as you suggest, 300GB-to-4TB end of the drives
compared to slower and 'real' data transfer rate on the
front 0-300GB end of the drives. HDDScan also has a
( destructive ) write test that puts the sector / LBA logical
block address number into each sector / LBA, and draws
a response graph - that might also accurately reveal or
corroborate the 'real' versus fake sections of the drives.

'Real' 4TB USB SSDs typically cost $200US or more.

See also the discussion in these GRC forums about SSDs
versus 'bit rot', where we share out experience of SSDs
NOT being appropriate for data storage, especially long
term storage, needing a periodic refresh just to maintain
readability, which is slow over USB, see
https://forums.grc.com/threads/ssd-...as-is-quite-real-a-spinrite-opportunity.2003/

My SAS-based servers also have SATA sockets, and SATA
drives are hot swappable, that is, we don't have to reboot,
but can toggle them on line and off line via software, in
Windows, Administrative Tools, Computer Management,
Disk Management, Action, Rescan Disks, right-click any
disk, select Offline or Online, and as such, a 4 TB or larger
SATA3 HDD should be comparatively affordable, as fast or
faster than USB, and promise more-reliable long term
storage with less maintenance - and even a USB HDD is
faster internally than USB2.

Tell us more, and let us know what you do.

 

[astrostar]

Thanks guys for all the really good information. The URL for the drive is
https://www.aliexpress.com/item/100...der_detail.order_detail_item.2.5da2f19co8NcnY
Maybe I should have realised the drives were probably dodgy due to the price - £27 each !
I reran the ValiDrive tests with the same results. Attached are some screenshots of the results.

Pics 1, 2 & 3 are prior to trying to copy the data
Pic 4 shows the Terracopy error screen
Pics 5, 6, 7, 8, 9 & 10 are showing the results after the failure.

Strangely, the first time I reran ValiDrive after the errors, it started to show a lot of red squares and then crashed. The second time I reran ValiDrive it showed only green squares.
I will run h2testw and HDDscan and see what results I get.

Many thanks for all your help.

 

[ColbyBouma]

Yep, that drive is definitely fake. It's unfortunate that it's able to get past ValiDrive. Definitely hang onto it so you can test ValiDrive v2

 

[ShadowMeow]

I got a fake 16TB drive off a friend, when I peeked inside I found a 8GB microSD, that works but has poor read/write speeds. The drive was £32, a real bargin for 16TB, shame it was fake. ValiDrive was what I used to check it was fake.

 

[astrostar]

Just ran HDDScan for a couple of hours and the program reported over 200.000 bad blocks !
The external enclosure supports 3Tb so I'm not sure if this is part of the problem ? I have another external enclosure on order and will rerun the tests with this when it arrives and let everyone know how it goes.

 

[peterblaise]

Yes, ValiDrive is a quick miracle, like a mini-SpinRite 6.1
Level 5 under Windows, but only for 576 'regions'.

SpinRite 6.1 under DOS, and a variety of programs under
Windows, such as HDDScan and h2testw, are so much
more thorough, checking millions of 'regions'.

Sorry you had to pay your way into the fake owner's club.

My entry fee was $20US for a '400GB' µSD that is just a
reprogrammed 32 GB.

That first 32 GB is surprisingly reliable, considering the
number of times I've done complete rewrite tests on it
by now, and it's still happy to await ValiDrive 2 in
development!

Someday, let us know if you can help develop ValiDrive 2,
or at least test those drives with it when development is
done.

 

[peterblaise]

I also have USB-to-SATA adapters that claim to support
only 3TB, and they work just fine with 20TB drives, no
problem - including initialization, partitioning, and
formatting.

I think the available drives determined the specification
limit, if any, when they created the instructions, and the
industry quickly surpassed it, but the underlying circuitry
had no true limits.

That being said, I don't run SpinRite 6.1 onto 20TB SATA
drives in USB adapters, I attach a SATA drive as SATA.

 

[astrostar]

Thanks for the info re: bit-rot I shall now look at conventional HDD's for archiving backups.
I am now awaiting Aliexpress to see if they do the decent thing and give me a full refund although I'm not holding my breath ! The seller initially offered a refund of £2.74 on a £27 drive !! At least members will know to avoid any drives being sold by the seller in the link I gave above.
I will keep you updated.

 

[peterblaise]

Using DiskPart and specific good-MB information from
h2testw, you can partition those as 300GB drives, and
use them for whatever you want.

My own 'got snookered' drive is surprising reliable and
robust in the 'real' area up front, surviving enormous
retesting and rewriting under ValiDrive 1 and SpinRite 6.1
development.

So even if you have to keep the 'fake' drives, they are
'good'-ish as 300GB each, and will be terrific test beds
for ValiDrive 2+ and SpinRite 7+ development testing -
no way should EITHER future program 'pass' those drives.

Note: ONLY SpinRite 6.1 Level 5 fails my fake drive, but
SpinRite 6.1 Level 3 and 4 pass, and Steve Gibson
suggests that SpinRite was designed to test drives that
don't lie, rather than to catch drives that do lie, so just be
aware that we may have to do more than 'trust' our prior
experience with our programs when we are assessing
these new-fangled constructs from cheaters.

Let us know what you do, keep in touch. Thanks.

 

[astrostar]

Just an update. Having retested the ssd drive by copying a number of smaller files I found that the maximum amount of data that could be written was around 103Gb before TeraCopy failed it's verification.
Being seriously miffed by now I decided to open the drive to see what it contained and after a struggle finally managed to open it up and surprise, surprise there's only one chip inside apart from (I'm assuming) the controller chip which has been painted over. I think the other chips must have been teleported out due to static as the seller stated that he had tested the drives and it must have been due to static in transit !!!!!!!!!!!

 

[astrostar]

For some reason the photographs didn't upload.

 

[astrostar]

Inside of drive

 

[rd369151]

For followups to ValiDrive, I run the long-running-with-files-copied type of tester called "MediaTester" from Github.