Jump to content











Photo
- - - - -

Is every bootable external USB Win XP married to the USB port it was created from? [Solved!]


  • Please log in to reply
11 replies to this topic

#1 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 25 June 2009 - 06:46 AM

Hi! I've just created a bootable external Win XP SP3 in a two-partition USB HDD (not a pendrive). I've used the ngine.de method (this is my first attempt at it), and it works beautifully. Yet, I found out it only works when I plug it exactly in the same USB port as the one I've used to create it. If I conect it to the same machine (an Asus eeePC 4G) using any of the other available two USB ports, it BSOD's.
While I can live with this perfectly, I was surprised, for I'd never read any mention to it. So, my questions are: Is a bootable external USB device created by ngine.de method really married to the USB port it was created from? Is there any workaround it? Is this phenomenon exclusively caused by the ngine.de method or would the USBoot method yield a similar result? Please do enlighten me about it.

#2 ktp

ktp

    Silver Member

  • Advanced user
  • 767 posts

Posted 25 June 2009 - 07:45 AM

From my experience, there is a method that works well : USBoot.
http://www.usboot.org/tiki-index.php

#3 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 25 June 2009 - 04:41 PM

Hi! I've just created a bootable external Win XP SP3 in a two-partition USB HDD (not a pendrive). I've used the ngine.de method (this is my first attempt at it), and it works beautifully. Yet, I found out it only works when I plug it exactly in the same USB port as the one I've used to create it. If I conect it to the same machine (an Asus eeePC 4G) using any of the other available two USB ports, it BSOD's.
While I can live with this perfectly, I was surprised, for I'd never read any mention to it. So, my questions are: Is a bootable external USB device created by ngine.de method really married to the USB port it was created from? Is there any workaround it? Is this phenomenon exclusively caused by the ngine.de method or would the USBoot method yield a similar result? Please do enlighten me about it.


It's the first time I heard about this kind of behaviour.

Can you re-build a setup on another port (one of the two that is now not working) and see if it BSOD's when connected on the one currently working (cross-checking, just in case it is not a specific EEE "one USB good/two bad" kind of specific issue/or different controller ;)).

BEFORE that, try checking USBbootwatcher service:
http://www.911cd.net...showtopic=22473

:lol:

jaclaz

#4 ktp

ktp

    Silver Member

  • Advanced user
  • 767 posts

Posted 25 June 2009 - 05:33 PM

BEFORE that, try checking USBbootwatcher service:
http://www.911cd.net...showtopic=22473


I like USBbootwatcher too since it looks simple and runs under XP and Vista/Windows 7. But unfortunately I got a case it failed under Vista/Windows 7 (Vista: after applying SP2).
There was my correspondence with Gerd (USBoot'author) about the the subject of USBbootwatcher, where he explained the problem with USBbootwatcher method:

http://www.usboot.or...e...c&forumId=1

For me I did not encounter yet a case where USBoot failed to work. Unfortunately it runs under XP only (and I believe the method of challenge/response codes is not user-friendly).

#5 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 25 June 2009 - 06:26 PM

I like USBbootwatcher too since it looks simple and runs under XP and Vista/Windows 7. But unfortunately I got a case it failed under Vista/Windows 7 (Vista: after applying SP2).

Well, since dencorso is using XP, that shouldn't be much of a problem. ;)

There was my correspondence with Gerd (USBoot'author) about the the subject of USBbootwatcher, where he explained the problem with USBbootwatcher method:

A small correction: that is some some correspondence where Gerd completely fails to explain:
http://www.merriam-w...tionary/explain
anything, all you can get from that is that Gerd states:
http://www.merriam-w...ictionary/state[2]
that his solution is better, which of course I consider NOT any kind of "news" or "info". ( I have rarely seen an Author failing to state the same about any of his own creations :lol:)


Unfortunately it runs under XP only (and I believe the method of challenge/response codes is not user-friendly).


The method of challenge/response code is in my view, together with the absolute lack of the single apps/services/scripts/whatever availability the BIGGEST ever drawback of the Gerd's app.

It effectively prevents experimenting anything different based on it.
It lessens motivation for actually experimenting (since it does work, and it's easy-peasy)
Thus, instead of contributing in spreading knowledge, it subtracts it from the public.

Let me be very clear about this, this is not at all a critic to the very good work by Gerd ;) , only a personal view on the IMHO unjustifiedly "secretness" surrounding it and all in all unfriendly way he chose to share the app.

I have similar feelings on the way chosen by other known and lesser known Authors, everything is about the difference way to interpret the word "free" (free as in "free beer" as opposed to "free as in freedom").

;)

jaclaz

#6 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 26 June 2009 - 04:50 AM

Well, in any case, however good the USBoot method may be, it requires one to install to an internal HDD, before transferring to the external one. And since my intention was to do everything without ever touching the (appallingly tiny) internal 4 GiB SSD present in the ASUS eeePC 4G, USBoot was not an option, so I went straight to the ngine.de method. And the result is perfectly satisfactory. The issue I reported did make me curious about whether it was widely known or not, since I had never read about anything like it. Then again, it's not much of a problem, because the actual external HDD I'm using needs two USB ports to run, so the cord lenght only permits connection to the side where there are two USB ports. The lone port on the other side is too far to be used. So one has just to remember to connect the two male USB connectors in the right arrangement, among two possibilities only. And, BTW, the BSOD is the infamous Stop 0x7B, and neither Dietmar's NTDETECT, nor the XP Embeded USB drivers help avoiding it. Win XP completes the first stage of boot and then BSOD's at the beginning of the second.

#7 Alakadebra

Alakadebra

    Newbie

  • Advanced user
  • 14 posts
  •  
    United States

Posted 26 June 2009 - 07:49 AM

We used to use BartPE with WinXP using the standard plugins for USB Thumb drive creation doing the boot iso through ram loading. This method appeared to work much more reliably for us but still had issues. It wasn't locked to usb ports that it was built on, we'd use it on several machines and different ports each time. We've since switched to VistaPE with Winbuilder and the USB thumb drive booting is even better now with Grub4Dos.

#8 was_jaclaz

was_jaclaz

    Finder

  • Advanced user
  • 7100 posts
  • Location:Gone in the mist
  •  
    Italy

Posted 26 June 2009 - 07:57 AM

We used to use BartPE with WinXP using the standard plugins for USB Thumb drive creation doing the boot iso through ram loading. This method appeared to work much more reliably for us but still had issues. It wasn't locked to usb ports that it was built on, we'd use it on several machines and different ports each time. We've since switched to VistaPE with Winbuilder and the USB thumb drive booting is even better now with Grub4Dos.


Hey, vkt2009, we have also used DOS, FREEDOS, ReactOS, several flavours of Linux and even QNX :lol:, but the present thread is about booting a "full" XP from USB, no more, no less. ;)

jaclaz

#9 Alakadebra

Alakadebra

    Newbie

  • Advanced user
  • 14 posts
  •  
    United States

Posted 26 June 2009 - 08:30 AM

Hey, vkt2009, we have also used DOS, FREEDOS, ReactOS, several flavours of Linux and even QNX :lol:, but the present thread is about booting a "full" XP from USB, no more, no less. ;)

jaclaz


Ahhhh sorry. That explains why you wanted the USB drive letter assignment stuff.

#10 online

online

    Silver Member

  • Advanced user
  • 767 posts

Posted 26 June 2009 - 12:40 PM

@dencorso

I would try two attempts (possibly mixing them too):
  • To boot from your USB HD with/without internal HD(s) connected within your machine.
  • To delete (after its backup) the "HKEY_LOCAL_MACHINE\SYSTEM\MountedDevices" key content.
Hoping it helps.

#11 karyonix

karyonix

    Frequent Member

  • Advanced user
  • 473 posts
  •  
    Thailand

Posted 27 June 2009 - 07:30 AM

ngine.de's instruction are good for same machine and USB port.
But is does not include CriticalDeviceDatabase, so Windows cannot boot on different hardware.

To move to different USB port, you should add some entries in
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\CriticalDeviceDatabase
like in Dietmar's tutorial.
  • dencorso likes this

#12 dencorso

dencorso

    Frequent Member

  • Advanced user
  • 142 posts
  •  
    Brazil

Posted 27 June 2009 - 04:08 PM

ngine.de's instruction are good for same machine and USB port.
But is does not include CriticalDeviceDatabase, so Windows cannot boot on different hardware.

To move to different USB port, you should add some entries in
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\CriticalDeviceDatabase
like in Dietmar's tutorial.

 

;) Way to go, karyonix! ;) You rock my life! ;) And welcome to Boot-Land!
I've created, by hand, these four keys in the registry of the live system, having customized the last two to my specific external USB HDD (here is a link to Dietmar's Tutorial 2, where this customization is explained in great details):

Windows Registry Editor Version 5.00

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\CriticalDeviceDatabase]

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\CriticalDeviceDatabase\usb#root_hub]
"Service"="usbhub"
"ClassGUID"="{36FC9E60-C465-11CF-8056-444553540000}"

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\CriticalDeviceDatabase\usb#root_hub20]
"ClassGUID"="{36FC9E60-C465-11CF-8056-444553540000}"
"Service"="usbhub"

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\CriticalDeviceDatabase\USB#Vid_058F&Pid_6390]
"Service"="usbstor"
"ClassGUID"="{36FC9E60-C465-11CF-8056-444553540000}"

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\CriticalDeviceDatabase\USBSTOR#DiskGeneric_USB_Disk________9.02]
"ClassGUID"="{4D36E967-E325-11CE-BFC1-08002BE10318}"
"Service"="disk"

...and the system booted like magic, from everyone of the three possible USB ports available in the machine!!! ;) Thank you a whole lot! ;)
And many thanks to everyone who contributed to this thread! ;)

Update: I added USBoot phase-0 and it eliminated the unsigned driver warning, in addition to letting me enable the page file in one of the USB HDD partitions. So, now I have an active, system managed, page file also. The downside is that every pendrive I insert is also seen as an unremovable disk, so there is no "eject" in the context menu. I can, however, stop it from the tray icon, as usual, so I don't see this as a problem. On the other hand, the challenge code is required to do this, so it's not for those viscerally opposed to it. IMHO, its freeware, all the same... and gets the job done. :lol: But I do understand the point-of-view of those who see it otherwise.






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users