New user, some initial obervations...

[crashbang]

Been using h2testw for maybe a decade now, but I find it's become almost unusable as flash memory cards have increased in size.

That's why I'm now trying validrive. I mean, it seems a obvious idea in principle - to test an evenly-distributed random sample of regions of the card, thus probably finding most fake/faulty cards without having to take all weekend running the test. Validrive being what turned up when I searched for that general idea. (No intention to dismiss the work involved in actually _implementing_ such a concept! Just saying the basic idea seems almost essential with cards and flash drives being the sizes that they are)

My object was to test a recent 1TB card I got that was having problems with (and as so often, can't tell if the problem is with the card, the device, some compatibility issue between the different elements involved, or the USB connection). But h2testw says it will take 9 _hours_ just to do the write part of the test!

Hence was hoping validrive would be better. It does give the card the all-green (and in a tiny fraction of the time that h2testw was going to take), though the 'worst' read and write speeds reported seem very slow, much slower than the 'average' speed it reports - am wondering if that signifies anything to worry about? Or is it 'normal'?

Also tested a couple of very old sd cards that I _know_ are fake/bad (because they never worked, and, if I remember correctly, h2testw eventually found them to be fake - they were the first sd cards I ever bought, when I didn't know any better, they seem to be 2GB claiming to be 32 and 64GB respectively). Am a bit confused/puzzled by the results I get.

One of them, the 64GB, validrive says is genuine (all green validated) but very, very slow - quite ridiculously so. The other one, the fake 32GB, comes out all green except for single red boxes at the start and end of the drive, but then reports (in the text part of the results) that there's no storage there at all!

 

[crashbang]

[Moderator warning: DO NOT download GRC utilities from ANY site but grc.com. I modified the URL presented by the user because that site is not safe to visit.]

Also, what's the difference between downloading validrive from the GRC site and downloading it from v a l i d r i v e . o r g or, for that matter, from techspot or softonic? I always get nervous when there are multiple download sites on offer, wondering if one or more of them might be fake/malware (as happened with 7zip not long ago) and then have to try and make a judgement as to which site is the 'genuine' one.

 

[PHolder]

First, and foremost: ValiDrive should only be downloaded from its page on grc.com at: https://www.grc.com/validrive.htm

Second, it's not magic, not impossible to fool, but then I doubt anything is. It's not taking drive throughput into its calculations of the validity of a drive. All it is doing it attempting to the use a small portion of the drive to verify that what it was requested to store it manages to succeed at retrieving later. While a good basic approach, it is not infallible, as "malicious" drive firmware has other ways to lie to a user's OS. As one example, the drive can operate as a FIFO and succeed at storing a small portion of the drives writes until they are bigger than the FIFO's size, and then throwing away the oldest data. If the size of the FIFO is bigger than what ValiDrive exercises, then it could be fooled.

As ever, the simplest advice is to buy quality devices from quality sellers (that offer a warranty and/or return/replacement policy) and to follow the old saw: if it's too good of a price to be true, it probably isn't actually a good deal. Lower than average priced devices got their lower price one of two ways: 1. they used lower quality flash that has failed testing at a more rigorous level and won't be as durable or as fast or as reliable; or 2. they're not offering what they claim, and it's lesser sized or less capable or just an outright scam.

 

[PHolder]

I removed all the posts speculating about any other version available. This is not helpful information because as I stated earlier: THE ONLY SAFE PLACE TO DOWNLOAD GRC APPLICATIONS IS AT GRC.COM. I don't want any speculation encouraging any user to do otherwise.

 

[peterblaise]

On the one hand, yes, always re-download from GRC BEFORE
continuing troubleshooting or asking for support, and run the
CURRENT version to confirm behavior.

On the other hand, good for folks to keep older GRC versions to
compare file hashes to confirm any GRC file found anywhere.

- - - - -

Now, to the surprises when scanning known fake drives ... PLEASE
share as much information about the drives as possible so @Steve
Gibson and others can try to figure out what is going on.

I occasionally get red boxes in ValiDrive results, sometimes
seemingly randomly, and wonder if I have a wonky USB connection,
so I retest on multiple computers and multiple USB ports.

And others have noticed flash drives that game the system and pass
ValiDrive, though failing on h2testw.

So even though h2testw may take a long time, it's probably an
important companion to ValiDrive, especially when trying to resolve
those wonky, otherwise unexplained red boxes.

So test overnight - but Windows is multitasking, so we can do other
things while testing one USB drive, right?

Here's an example of a fake drive working hard to dance with
ValiDrive, but unable to keep track of what data to return, so it's
different every time, except for the initial 'real' storage:

I wonder what are the results from and causes of:

"... seem to be 2GB claiming to be 32 and 64GB respectively). Am a ​

bit confused/puzzled by the results I get. One of them, the 64GB, ​

validrive says is genuine (all green validated) but very, very slow - ​

quite ridiculously so. The other one, the fake 32GB, comes out all ​

green except for single red boxes at the start and end of the drive, ​

but then reports (in the text part of the results) that there's no ​

storage there at all! ..."​

​

Do we have new flash drives awaiting a 'smarter' ValiDrive 2"

 

[peterblaise]

"... the fake 32GB, comes out all green except for single red boxes at the
start and end of the drive, but then reports (in the text part of the
results) that there's no storage there at all! ..."​

The 'total storage' report is based on good storage up to the first
failure.

So if the first box is 'dead', then ValiDrive will consider that to be the
end of the drive, so to speak.

Supposedly 'good' storage after any failed storage is not considered.

Here's a fake 400 GB uSD based on a 'good' 32 GB uSD:

Note that ValiDrive does mention the higest valid region up to
277 GB on this test ( different every time ), but the validated drive size
remains 31.4 GB nevertheless.

- - - - -

Yes, writes are way slower than reads within ValiDrive's techniques.

If we run ValiDrive on ALL our USB storage on ALL our USB sockets,
then we will become familiar with performance anomolies.

However, ValiDrive is NOT a speed benchmark, and does NOT read
and write the way Windows does.