This is the mail archive of the cygwin mailing list for the Cygwin project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: ntfsclone for cygwin 1.7 (issues pointing to shadow copies)

On Jul 27 18:31, Larry Hall (Cygwin) wrote:
> On 07/27/2009 05:48 PM, Rob Bosch wrote:
>> The issue appears to be with ntfsclone in that it must point to
>> something  in /proc/partitions. It will not work with any /cygdrive/X (used
>> dosdev.c from Corinna and typical vshadow to expose as a drive with no luck). I found
>> this out by commenting out the read-only check and rebuilding
>> ntfsclone/ntfsinfo. The code apparently uses a win32_io.c file when building
>> on cygwin. This file creates its own device mappings (e.g. /dev/hda1 instead
>> of using the existing /dev/sda1). I guess there would be a way to modify it
>> to use cygwin devices but I don't know how.
>> If shadow copies could be presented as block devices to this library
>> then  it would open up a lot of possibilities for creating block-level clones of
>> partitions without having to boot to a live CD. I guess it is dd piped
>> through gzip for now. It sure is a waste when you only have 50GB on a 500GB
>> drive since dd processes the entire drive.
>> Does anyone have anything else to add?  I'm guessing not...
> Not much in the way of specific help, no, but I can make two observations:
>   1. Using win32_io.c under Cygwin is almost certainly the wrong thing,
>      since you'll then be using Win32isms in an environment built to expect
>      POSIX mechanisms.
>   2. It looks like /proc/partitions isn't populated, so if /proc/partitions
>      is the only way that ntfsclone expects to be able to find what it
>      wants, then it will be disappointed until a patch is supplied to
>      support this.

/proc/partitions is populated, but only with actual disks.  The device
mapping for disks is a mapping from the internal NT namespace objects
\Device\HarddiskX\PartitionY to the POSIX-like /dev/sdXY.  Since shadow
copies of a disk don't match this mapping, you're out of luck.  Maybe
we can add a mapping in a later Cygwin release and I added it to my
TODO list, but this is definitely no fodder for 1.7.1.


Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader          cygwin AT cygwin DOT com
Red Hat

Problem reports:
Unsubscribe info:

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]