[sorry, forgot to send the mail to the list too]

Hi,

> It might not work so well.  Unionfs works based on files, it's unions 
> file system namespaces into a single namespace, but the restrictions of 
> each individual file system are still present.
> 
> If you have 3 4 TB file systems, the largest file size that could be is 
> 4 TB.  The other issue is that the you write to the topmost file system 
> (unless modifying an already existing file system) so it wouldn't use 
> the file system in the most efficient manner.  You can get around this 
> by making a daemon that rotates the unioned layers around (if that suits 
> your purposes).

our files won't be larger than 100-200 GB (AFAIK). If I understand
you right it would be a problem if 3.9 TB of the first fs are already
occupied and a new file with 200 GB should be copied to that unionfs?

The rotation of the layer is somethin I don't quite understand, I'm
very new to unionfs. But I found a fakeraid path for unionfs.

Unionfs File Spreading
http://www.fsl.cs.sunysb.edu/pipermail/unionfs/2006-January/001600.html

Ralf
_______________________________________________
unionfs mailing list: http://unionfs.filesystems.org/
[email protected]
http://www.fsl.cs.sunysb.edu/mailman/listinfo/unionfs

Reply via email to