storage considerations

[Serge E. Hallyn]

Quoting Dietmar Maurer (dietmar at proxmox.com):
> > >>> Dietmar Maurer wrote:
> > >>>
> > >>>> Hi Daniel,
> > >>>>
> > >>>> I think we should have several options for the root storage:
> > >>>>
> > >>>> 1.) simply use the host filesystem (like Openvz)
> > >>>>
> > >>>> 	- special quota support is needed (simfs?)
> > >>>>
> > >>>> 	- quota support depends on ext3 fs, so this only works for
> > >>>> 	  local attached storage (does not work on NFS)
> > >>>>
> > >>>> 	- LVM snapshots are slow, because the snapshot includes all
> > >>>> container
> > >>>> 	  on that filesystem.
> > >>>>
> > >> I'm looking forward to the btrfs, tux3, and nilfs snapshotting
> > >> capabilities.
> > >>
> > > Interesting, especially the nilfs.
> > > The COW file systems is missing in the linux kernel. Do you think
> the
> > > btrfs has a chance to go to mainline ?
> > 
> > We've been using NILFS (v1) for 2 years already with Zap, and it works
> > quite well. We had to make slight (trivial) changes to be able to
> > control it nicely for c/r purposes.
> 
> NILFS, BTRFS, ... maybe provide solution for snapshots, but quota
> support
> is still missing (we need quotas on subdirectories)?

Haven't thought it through 100%, but does it not suffice to be able
to specify quotas per user-namespace?  It won't help with NFS...

> And a FS independent solution would be great, so that it works on top of
> NFS too. 
> 
> So UnionFS looks good, but it does not provide real snapshots (using 
> base + incremental change looks clumsy for our purpose). I am also not
> sure
> if quota works on UnionFS?
> 
> - Dietmar