Store once and all reference
I was reading an article about another internet company coming up with an
online storage solution:
http://blogs.zdnet.com/Google/
And it reminded me of the times when the network storage at my company would run
out of space. I would start looking around in user directories and notice that
300 users would all have the same
ManualsAtoZ.pdf file.
At 800 kb it took up 234 mb of space. So I would remove the file from every
users folder, put it in a directory accessible by all, and send a short "ManualsAtoZ.pdf
has been removed from your Home
drive and moved to the shared drive at M:\Manuals\ManualsAtoZ.pdf" email to all
300 users.
Does anyone recall how eDonkey would be able to find similar files because they
would have the same MD5 hashes? What if the storage solution of the future had
all duplicate files only saved once, and
you had the equivalent of a shortcut to that file in your Home drive (or GDrive
as Google may be calling it)?
To the user it would be completedly transparent, but on the server itself, it
would only save that file once, keeping the storage sizes small, and every user
could think they had that file (and if
they "deleted it", it would just remove the reference from their Gdrive, not
everyones).
How say you?