Hi All
One of the things I discovered a long time ago is that there was no
proper error checking for drives on the USB bus. I have had old
computers where I could not transfer data using the USB.
Something similar could be happening here.
Create a large file and take an SHA of the file. Send both files to
the disc. Take an SHA of the file on the disc and compare it to the
first SHA. If there is a problem with data transfer on the USB bus
then the two SHAs will be different.
John Eayrs
On Fri, 5 Jan 2018 16:08:22 -0000
Rob Malpass via Hampshire <hampshire@???> wrote:
> Hi all
>
>
>
> I can't explain this - perhaps someone else can. I have a 2TB 3.5"
> ext4 formatted internal drive (bought only in December) which was
> reporting errors yesterday. At the time, it was connected via USB
> Icybox JBOD and threw out more "short read" errors than I could
> count. I left it running
>
>
>
> e2fsck -y /dev/blablabla
>
>
>
> overnight and it was still reporting errors this morning. Getting
> fed up (and not wanting to write the unit off), I did
>
>
>
> mkfs /dev/blablabla
>
>
>
> and when I fired up rsync again - same type of IO errors. At this
> point I wrote the drive off and used a spare for my purposes.
>
>
>
> Curious to see if I could get any more information from the failing
> drive, I then moved it into a USB docking station on a different
> machine. Running e2fsck again and I got a clean filesystem (no new
> formatting or anything).
>
>
>
> Worried that this meant the drive was fine and potentially that
> particular bay in the (4 bay) Icybox might be the culprit, I moved
> the rogue drive back into the JBOD (same bay) and guess what - clean
> bill of health from e2fsck.
>
>
>
> So in short I have the same drive reporting errors, reformatted
> reporting errors, physically moved clean, then physically moved back
> clean. I've never been too hot on the rather low level way Linux
> handles disks - but I do want to know if the effing thing is good to
> use or not.
>
>
>
> Is a clean e2fsck result good enough? If so, were the hundreds of
> errors it was chucking out safely ignorable? Have I missed
> anything obvious?
>
>
>
> Cheers
>
> Rob
>
--
Please post to: Hampshire@???
Web Interface:
https://mailman.lug.org.uk/mailman/listinfo/hampshire
LUG URL:
http://www.hantslug.org.uk
--------------------------------------------------------------
