[sf-lug] Trying to determine the actual size of a USB drive

[Jeff Bragg]

On Mon, Aug 29, 2011 at 8:32 PM, Jeff Bragg <jackofnotrades at gmail.com>wrote:

> On Mon, Aug 29, 2011 at 8:06 PM, John Magolske <b79net at gmail.com> wrote:
>
>> I have an OCZ USB 8GB Dual Channel Rally2 Pen Drive, and for some
>> reason df seems to be telling me it has more than 8GB capacity:
>>
>>     ~ % df -h
>>    Filesystem            Size  Used Avail Use% Mounted on
>>    [...]
>>    /dev/sdb1              15G  7.3G  6.7G  53% /media
>>
>> I'd like to determine with certainty whether this is an 8GB or 16GB
>> drive. Someone suggested copying a file larger than 8GB to the drive
>> and comparing the copied file to the original...maybe with CMP(1) or
>> using checksums. I'm wondering if there might be another way, as I'd
>> rather not erase everything currently on the drive. I'd also like to
>> avoid adding more write cycles to its flash storage chip.
>>
>
>
> I've never worked with dual-channel USB, so I might be off-base here, but
> I'm wondering if it's confusing the USB driver.  If dmesg doesn't yield
> anything enlightening, there's a way to indirectly test the drive's capacity
> (sort of).  Rather than trying to fill up the drive, you could copy what's
> on it off into a directory on the same host, then find out how big the host
> believes that is.  If it matches the 6.7GB expected, I would tend to assume
> you have a larger drive than you believed.  If, on the other hand, it's
> about half of that, I would tend to believe the dual-channel is causing
> everything to be doubled, and the drive is 8GB.  The really confusing result
> would be having ~6.7GB as expected, but only actually having ~1.3GB left
> (which I can't think of a way to discover other than trying to copy a file
> slightly bigger than that to it and see if it fails).
>

Oops, used the wrong #'s there.  Substitute 7.3GB for 6.7GB and 0.7GB for
1.3GB where relevant.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://linuxmafia.com/pipermail/sf-lug/attachments/20110829/fc21983e/attachment.html>