Hi Everyone,
I'm back!
First thing I want to make mention of is that using "pci=nomsi" disabled all of the USB controllers/connections. I had a few uncomfortable minutes with that because I forgot to "update-grub" again after I reset /etc/default/grub to its original state. I only just tried this before starting to compose this post.
Ok then, over the last few days, I've discovered some annoying things going on in both the Linux and Windows environments.
In Windows, I was able to repeat the problem of a USB 3.0 Flash Drive not being recognised as a USB 3.0 device in a USB 3.0 port.
This could occur after the USB Drive was plugged into a USB 2.0 port and whether, or not, files were copied to the drive.
Prior to this I had thought that UNetbootin had corrupted the drive in some way. So, I started looking for solutions to reset a USB Flash drive so that it was recognised as a USB 3.0 Device. Hence, I discovered the following page.
http://superuser.com/questions/700254/u ... -windows-8
So, after I connected the drive to a USB 2.0 port, demounted and removed it and then connected it to a USB 3.0 port, 3 times, Windows recognised the drive as being a USB 3.0 device again. Sometimes, just demounting and removing from the USB 3.0 port and connecting it again to the USB 3.0 port, without any use in a USB 2.0 port, has the same effect.
Also, if the Flash drive is in a USB 3.0 port, when Windows is started, it is recognised as a USB 3.0 device.
So, I thought I would try this in the Linux environment but no success with switching between a USB 2.0 and USB 3.0 port as in Windows. Sometimes I get a
"not running at top speed; connect to a high speed hub" in the output from dmesg when it is connected to a USB 2.0 port.
The only way make to make sure it is recognised as a USB 3.0 device, in Linux, was to have it in a USB 3.0 port when the Operating System was started. Just once, today, I managed to demount, remove and reconnect the drive to the same USB 3.0 port and it was recognised again as a USB 3.0 device (must have been a fluke).
I also have a USB 3.0 Hard Drive Enclosure that is properly recognised in the USB 3.0 ports without the above problems. Its USB cable has connectors for both USB 3.0 and USB 2.0 ports but the USB 2.0 connection doesn't appear to be working even though I can hear the drive being powered up.
So, I discussed this with one of the technicians at the office where I bought the Laptop.
They had a Toshiba Laptop with Windows 8.0 and a Desktop with Windows 7.
We could repeat the problem on the Toshiba but not on the Desktop with Windows 7.
He then repeated the same testing with a Kingston DataTraveler 100 G3 16GB Flash Drive which he said he had found to be the most reliable USB Flash Drive he had used. It was formatted as NTFS. It functioned correctly with both the Toshiba and the Desktop.
So, I bought a Kingston DataTraveler 100 G3 8GB Flash Drive.
In the Linux Environment, this new Drive was also not recognised as using the USB 3.0 protocol when connected to a USB 3.0 port after the Linux Desktop was activated.
If the Drive was connected to a USB 3.0 port before the Laptop was booted, then, Linux recognised the Drive as being a USB 3.0 (SuperSpeed) device.
Code: Select all
[ 3.171083] usb 4-4: new SuperSpeed USB device number 2 using xhci_hcd
[ 3.281978] usb 4-4: New USB device found, idVendor=0951, idProduct=1666
[ 3.281986] usb 4-4: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[ 3.281990] usb 4-4: Product: DataTraveler 3.0
[ 3.281993] usb 4-4: Manufacturer: Kingston
[ 3.281996] usb 4-4: SerialNumber: 08606E6D418ABE613717B8DB
[ 3.283307] usb-storage 4-4:1.0: USB Mass Storage device detected
[ 3.283419] scsi7 : usb-storage 4-4:1.0
If the Drive was demounted and removed from the USB 3.0 port and then connected again to the same USB 3.0 port whilst Linux was operational then the Drive was not recognised as using the USB 3.0 protocol as shown below. It was recognised as being a USB 2.0 (High-Speed) device.
Code: Select all
[12249.810725] usb 4-4: USB disconnect, device number 2
[12249.812086] usb 4-4: Set SEL for device-initiated U1 failed.
[12249.812096] usb 4-4: Set SEL for device-initiated U2 failed.
[12264.195528] usb 3-4: new high-speed USB device number 8 using xhci_hcd
[12264.422082] usb 3-4: New USB device found, idVendor=0951, idProduct=1666
[12264.422090] usb 3-4: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[12264.422094] usb 3-4: Product: DataTraveler 3.0
[12264.422097] usb 3-4: Manufacturer: Kingston
[12264.422101] usb 3-4: SerialNumber: 08606E6D418ABE613717B8DB
[12264.422681] usb-storage 3-4:1.0: USB Mass Storage device detected
[12264.423098] scsi13 : usb-storage 3-4:1.0
In the Windows 8.1 Pro Environment, initially, everything appeared fine.
The Drive was recognised as using the USB 3.0 protocol when connected to a USB 3.0 port after the Desktop was activated. The interface indicated that it was
"Connected to USB 3.0".
On moving the Drive to a USB 2.0 port, the interface indicated that the
"Device can perform faster when connected to USB 3.0".
Moving the Drive back to a USB 3.0 port, the interface indicated that it was
"Connected to USB 3.0".
Then, I wondered what would happen if I copied a file to the drive whilst it was in the USB 2.0 port.
So, I copied a file and after demounting and removing the drive from the USB 2.0 port, I restarted Windows 8.1 Pro.
When I connected the drive to the USB 3.0 port again, the interface indicated that the
"Device can perform faster when connected to USB 3.0".
So, I tried the solution of swapping between the USB 3.0 and USB 2.0 ports. I did this 4 times without success.
Then, for nothing better to do, I demounted and removed the drive from the USB 3.0 port and then connected again to the same USB 3.0 port and once again the interface indicated that it was
"Connected to USB 3.0". Strange!
In both Operating System environments the DT100G3 has significantly better Read performance than the DTIG4 but the Writing performance is very poor, in the Linux environment, and only half of that of the DTIG4.
However, in the Linux environment, the Sequential Read performance is only a little more than half that of the Windows environment. I tested with FAT32, NTFS, ext2, ext3 and ext4 file systems and compared the results against those of CrystalDiskMark and ATTO Disk Benchmark in Windows 8.1 Pro but I'm not sure if I'm comparing Apples with Apples or Apples with Pears.
Interestingly, the Write performance was best with NTFS, ext2 and ext3 in the Linux environment according to the following tests but wasn't corroborated by the Gnome Disk Utility benchmark. Also, I can't say how valid the following tests are. Hopefully, someone with more experience can enlighten us all.
Code: Select all
NTFS
dd count=100 bs=1M if=/dev/urandom of=/media/mtbvfr/KINGSTON/test
100+0 records in
100+0 records out
104857600 bytes (105 MB) copied, 8.20797 s, 12.8 MB/s
dd count=100 bs=1M if=/dev/urandom of=/media/mtbvfr/KINGSTON/test oflag=sync
100+0 records in
100+0 records out
104857600 bytes (105 MB) copied, 7.96686 s, 13.2 MB/s
ext3
dd count=100 bs=1M if=/dev/urandom of=/media/mtbvfr/KINGSTON/test
100+0 records in
100+0 records out
104857600 bytes (105 MB) copied, 6.92441 s, 15.1 MB/s
dd count=100 bs=1M if=/dev/urandom of=/media/mtbvfr/KINGSTON/test oflag=sync
100+0 records in
100+0 records out
104857600 bytes (105 MB) copied, 21.6813 s, 4.8 MB/s
ext2
dd count=100 bs=1M if=/dev/urandom of=/media/mtbvfr/KINGSTON/test
100+0 records in
100+0 records out
104857600 bytes (105 MB) copied, 6.82789 s, 15.4 MB/s
dd count=100 bs=1M if=/dev/urandom of=/media/mtbvfr/KINGSTON/test oflag=sync
100+0 records in
100+0 records out
104857600 bytes (105 MB) copied, 44.3509 s, 2.4 MB/s
So, now we are back where I started from. Why is the Sequential Read performance substantially less than in the Windows 8.1 Pro Environment?
It would be helpful to know if anyone else can repeat, or not repeat, these problems, that I have encountered, with the same, or different, brands and models of USB Flash Drives with the same operating systems.
Having to restart an operating system in order to have a USB 3.0 Flash Drive recognised as such isn't very convenient.
Thanks, Michael.