Comparing existence of massive number of files

From: J.C. Wren <jcwren_at_jcwren.com>
Date: Tue Mar 30 08:29:33 2004

    Many versions of diff support diff'ing contents of directories,
usually by just specifying the two directories as the targets. "diff -r
pathToDir1 pathToDir2". This will work on almost any filesystem type
that can be mounted under Linux. And if you don't have Linux installed,
consider using a live CD, perhaps Knoppix or Lindows.

    You should also be able to do this with Cygwins diff. Most Cygwin
utils will run with just the cygwin1.dll and the .exe file. If you
don't have Cygwin installed (and I believe someone mentioned this
earlier), you can install a comprehensive chain of *nix tools with a
minimum footprint. Just don't install GCC :) It's very handy when you
want command line tools that aren't as readily available under Windows,
or you're a *nix hack who switches between Windows and *nix.

    --jc

Paul Koning wrote:

>>>>>>"Richard" == Richard A Cini <rcini_at_optonline.net> writes:
>>>>>>
>>>>>>
>
> Richard> Hello, all: I just moved most of my collection of on-line
> Richard> data to a new server here at home. I want to verify that all
> Richard> files have been copied and I didn't miss any
> Richard> directories/subdirectories. Basically I did it through a
> Richard> drag-and-drop across the network.
>
> Richard> What's the best way to do this? The OS is Windows NT
> Richard> (Server) and I'm looking at about 22gb of files of various
> Richard> types -- from music to source code.
>
> Richard> Any ideas? I still have the original server on standby for
> Richard> this verification before I wipe it clean for sale.
>
>Boot Linux, ls -lR on the Windows file system at each end, diff the
>listings.
>
>If you're paranoid, do a cksum on the files instead, and compare those
>listings. That will verify not just the existence but the content as
>well.
>
> paul
>
>
>
Received on Tue Mar 30 2004 - 08:29:33 BST