On Jun 27, 2011, at 3:39 PM, Alan Stern wrote: > On Mon, 27 Jun 2011, M. Steven Ginn wrote: > >> >> On Jun 27, 2011, at 1:49 PM, Alan Stern wrote: >> >>> On Mon, 27 Jun 2011, M. Steven Ginn wrote: >>> >>>> ok attached is the output from usbmon (assuming I did it correctly) >>>> >>>> please let me know if I didn't and need to rerun it. I followed the >>>> instructions using cat to capture the output to 1.mon.out and then >>>> plugged in my USB 3.0 device >>> >>> The file looks okay. And it confirms what I suspected: The Addonics >>> USB interface is buggy. It tells the computer that five drives >>> are attached when there really are only four. >>> >> >> so is this a firm ware issue with the Addonics? (i.e. something Addonics should fix?) > > Yes. They'll probably tell you that there's nothing wrong because it > works under Windows. Don't believe them. I'm sure > >>> The direct ESATA interface doesn't have this bug. That would explain >>> why it works when you hook it up that way and not when you hook it up >>> over USB. >>> >>> Also, this appears to have no connection at all with USB-3. If you >>> plug the adapter into a USB-2 port, you should see the same kind of >>> failure. Try it out. >>> >>> On the other hand, failure to work with the fifth drive shouldn't >>> prevent the first four from working. Your usbmon trace didn't show >>> what happens when you try to mount one of the drives. >> >> Is there any point in trying to determine this if it's an Addonics firmware issue? > > Of course there is. If you can still use the device in spite of the > bad firmware, you'll be in good shape. ok that would be nice … the concept seems very attractive … increase throughput from esata II to USB 3.0 5gps here is a link to their website talking about the use of a different kind of USB port called esata USB Hybrid (EUHP) that is being used in the Addonics device. Any chance the implementation of power on the esata side could be causing the issues? > >> I can't mount any of the drives because they don't show up in either df or gparted when connected via the Addonics port. > > Do the drives show up under /dev? I haven't checked … I just assumed that if they don't show up with df or gparted, they didn't exist as far as the system was concerned. > >>> One last thing: It would simplify the usbmon output if you trace only >>> the bus to which the device is attached. In this case, that means >>> copying the 4u file instead of the 0u file, since the drive was >>> attached >>> to bus 4. >> >> Should I re-run this with the 4u file, given the above? > > Yes. And don't stop too soon. The file you posted before ended in the > middle of a command. Remember, each time the Addonics gets > seriously messed up, the kernel waits 30 seconds before cancelling the > command and resetting the device. That happened twice in the usbmon > trace, but you stopped recording after only 34 seconds! > > (You may have to wait a _long_ time before the other drives begin > working, though. There could be lots of errors with lots of retries, > and each will require 30 seconds.) ok I will run it on 4u and wait for a while .. maybe run Sarah's pipe of kern.log in another terminal to see when it finally gives up and I remount under esata. > > Alan Stern > Thanks for your patience. Steve sginn@xxxxxxx -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
