Boston Linux & Unix (BLU) Home | Calendar | Mail Lists | List Archives | Desktop SIG | Hardware Hacking SIG
Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings
Linux Cafe | Meeting Notes | Blog | Linux Links | Bling | About BLU

BLU Discuss list archive


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Discuss] restoring Windows on different hardware



On Thu, 27 Dec 2012 21:31:01 -0500
Tom Metro <tmetro+blu at gmail.com> wrote:

> https://www.virtualbox.org/wiki/Migrate_Windows
> 
>   For reasons we don't understand, Windows memorizes which IDE/ATA
>   controller it was installed on and fails to boot in case the
>   controller changes. ... The solution here is to perform several

Linux does something similar. Take your modern Linux workstation,
reboot, start the BIOS/EFI configuration, and change the SATA
controller mode from AHCI to Legacy. Now reboot the system and
watch the kernel panic when it can't find the root file system.

Your issues with XP arose from the fact that Windows XP doesn't ship
with SATA/AHCI drivers. You'll have the same problem with a ten year
old Linux kernel that has no SATA/AHCI drivers.

 
> Consider all the Linux live CDs that boot on just about anything. I'm
> willing to bet they more closely resemble your installed Linux than a
> Windows install CD environment matches your installed Windows OS.

You'd lose that bet. Roughly half of the Live CDs that I've played
around with won't boot correctly or at all on a Ivy Bridge box (note:
may be Ivy Bridge configured for Intel SRT). It's the dynamic device
enumeration "feature" in the Linux kernel. The loader passes a root
file system device to the kernel. The kernel ignores the passed device
ID and figures it out for itself, gets the wrong answer, can't find the
root file system, and panics.

You also get inconsistent behavior if you boot a Linux live CD as a
legacy device vs. as an EFI device. Inconsistencies include boot
failures due to device enumeration cock-ups and incorrect numbers of
CPU cores being detected. Two specific examples:

The Kaspersky Rescue CD I mentioned a couple of weeks ago. It won't
boot from USB at all on my notebook (kernel panic) but it will boot
from CD-R.

The most recent Clonezilla live CDs. Will boot either USB or CD. If the
USB is booted as a legacy device then everything works correctly. If
the USB is booted as an EFI device then it sees one core with one
thread instead of four cores with 8 threads.

I've yet to see a Windows PE (Windows Pre-installation Environment,
the Windows equivalent to a Linux live CD) disc that fails to start
properly. Admittedly, I've seen fewer WinPE and BartPE discs than Linux
live CDs.


> One thing you definitely wont see after booting up a Linux system
> after the drive is moved to new hardware is a message that your
> system needs to be activated. :-)

This is a support licensing issue (not a DRM issue) that's usually
resolved with a quick telephone call. You'll have similar issues with
Linux support contracts (RHEL, Oracle, etc).

-- 
Rich P.



BLU is a member of BostonUserGroups
BLU is a member of BostonUserGroups
We also thank MIT for the use of their facilities.

Valid HTML 4.01! Valid CSS!



Boston Linux & Unix / webmaster@blu.org