Hi,

first of all thanks for the input. I guess if you create a hash before
encryption and only store the encrypted file if the hash doesn't exist
the file/chunk has to be decrypted/encrypted by all users in the same
way. Otherwise they would have to store your user specific encryption
result. At least thats what I would think. There is actually an
article on how they do the encryption: 
http://support.crashplan.com/doku.php/articles/encryption_key.
Maybe there is still enough room for deduplication after the files
have been encrypted. Could this be?

I hope I'm not bothering anyone with the questions. It just seems
interesting to me since I always thought this was some kind of a
tradeoff decision.


On Mar 6, 8:46 am, Karsten Silz <[email protected]> wrote:
> On 5 Mrz., 17:29, Sebastian Himberger
>
> <[email protected]> wrote:
> > What I don't understand: If the encryption happens on the client side
> > and is done right this shouldn't be possible, right? Or am I missing
> > something? I'm certainly no expert on this.
>
> If I was to deduplicate on the client, I would calculate an SHA1 or
> MD5 hash over a file and send it, together with the file name and the
> file size, to the server.  If a file with the exact same three
> attributes already exists on the server, it's the same and you don't
> need to send the file over.  Otherwise, encrypt it and send it over.
>
> Calculating SHA1 or MD5 hash is pretty cheap, so it's worth it.  This
> works well for program or OS files, but fails for something like
> iTunes music (though it doesn't have DRM in it anymore, it's
> supposedly has your email address and some other metadata in it, so
> one million Lady Gaga songs were all different).

-- 
You received this message because you are subscribed to the Google Groups "The 
Java Posse" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/javaposse?hl=en.

Reply via email to