[OpenIndiana-discuss] ZFS Snapshot question

[Mark Creamer]

Thanks Jan, that sounds easy enough. I'll try it with a non-critical one
that I use for testing but your explanation looks like it will be more
straightforward than I expected. I appreciate the help.

On Thu, Aug 2, 2012 at 4:59 PM, Jan Owoc <jsowoc at gmail.com> wrote:

> Hi Mark,
>
> On Thu, Aug 2, 2012 at 2:53 PM, Mark Creamer <whitetr6 at gmail.com> wrote:
> > What I should have done (I think) is set up the zfs file system for each
> VM
> > at the level below ../vmimages, that is:  "zfs create
> > datastore/vmimages/server1". That way, auto-snapshot would be creating
> > snapshots for each VM, rather than for the entire ../vmimages file system
> > each time.
> >
> > So, is there an easy way to make a separate zfs file system for each of
> the
> > existing directories below ../vmimages? I am able to take the VM guests
> > down to do this as needed.
>
> Yes, zfs filesystems can be nested as many times as you like. On a
> single pool, you could have "/", then "/home", then "/home/mark", and
> then even create subfilesystems for different kinds of data
> (compressible data, data that needs copies=2, etc.).
>
> Having said that, I believe they need to be mounted in empty
> directories, so you'd have to do:
> # mv datastore/vmimages/server1 datastore/vmimages/server1.bak
> # zfs create -o [options here] datastore/vmimages/server1
> # cp -a datastore/vmimages/server1.bak/* datastore/vmimages/server1
> # rm -r datastore/vmimages/server1.bak
>
> Jan
>
> _______________________________________________
> OpenIndiana-discuss mailing list
> OpenIndiana-discuss at openindiana.org
> http://openindiana.org/mailman/listinfo/openindiana-discuss
>



-- 
Mark