zfs discuss - Apr 2010 - Is file cloning anywhere on ZFS roadmap

Hi,
We are currently using NetApp file clone option to clone multiple VMs on our FS.

ZFS dedup feature is great storage space wise but when we need to clone allot of
VMs it just takes allot of time.

Is there a way (or a planned way) to clone a file without going through the
process of actually copying the blocks, but just duplicating its meta data like
NetApp does?

TIA,
-- Schachar
-- 
This message posted from opensolaris.org

On 21/04/2010 07:41, Schachar Levin wrote:
> Hi,
> We are currently using NetApp file clone option to clone multiple VMs on
our FS.
>
> ZFS dedup feature is great storage space wise but when we need to clone
allot of VMs it just takes allot of time.
>
> Is there a way (or a planned way) to clone a file without going through the
process of actually copying the blocks, but just duplicating its meta data like
NetApp does?
>
>    
I don''t know about file cloning but why not put each VM on top of a
zvol
- then you can clone a zvol. ?


-- 
Robert Milkowski
http://milek.blogspot.com

You''d run out of LUN IDs on the VMware side pretty quickly (255 from
what I remember).

It''s also not really VMware best practice for block.

-marc

On 4/21/10, Robert Milkowski <milek at task.gda.pl>
wrote:
> On 21/04/2010 07:41, Schachar Levin wrote:
>> Hi,
>> We are currently using NetApp file clone option to clone multiple VMs
on
>> our FS.
>>
>> ZFS dedup feature is great storage space wise but when we need to clone
>> allot of VMs it just takes allot of time.
>>
>> Is there a way (or a planned way) to clone a file without going through
>> the process of actually copying the blocks, but just duplicating its
meta
>> data like NetApp does?
>>
>>
>
> I don''t know about file cloning but why not put each VM on top of
a zvol
> - then you can clone a zvol. ?
>
>
> --
> Robert Milkowski
> http://milek.blogspot.com
>
> _______________________________________________
> zfs-discuss mailing list
> zfs-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
>
-- 
Sent from my mobile device

On Wed, April 21, 2010 02:41, Schachar Levin wrote:
> Hi,
> We are currently using NetApp file clone option to clone multiple VMs on
> our FS.
>
> ZFS dedup feature is great storage space wise but when we need to clone
> allot of VMs it just takes allot of time.
>
> Is there a way (or a planned way) to clone a file without going through
> the process of actually copying the blocks, but just duplicating its meta
> data like NetApp does?
You mean like the ZFS "clone" command?

      zfs clone [-p] [-o property=value] ... snapshot filesystem|volume
>From zfs(1M):
[...]
 Clones
     A clone is a writable volume or file  system  whose  initial
     contents are the same as another dataset. As with snapshots,
     creating a clone is nearly instantaneous, and initially con-
     sumes no additional space.

     Clones can only be created from a snapshot. When a  snapshot
     is  cloned,  it  creates  an implicit dependency between the
     parent and child. Even though the clone is created somewhere
     else  in the dataset hierarchy, the original snapshot cannot
     be destroyed as long as a clone exists. The origin  property
     exposes  this  dependency, and the destroy command lists any
     such dependencies, if they exist.
[...]

So you''d have something like the following:

   # zfs create /mypool/machine00
   [ Create VMDK stuff. ]
   # zfs snapshot /mypool/machine00 at template
   # for i in 01 02 03 04 05 \
     do \
         zfs mypool/machine00 at template mypool/machine${i} \
     done
   #

This should work for both regular file systems (shared via NFS) and zvols
(via iSCSI); at $WORK we''ve used the latter to clone LDoms vdisks.

f I could mount unlimited NFS mounts on ESX it would be good for me, but
there is a 64 nfs datastore limit.

Right now if I copy a big file it can take several minutes. I already tried
having one VM per zfs file system, and snapshot & clone works great, as long
as I stay below 64 NFS mounts.
Note that VMWARE doesn''t support NFS v4 or else doing something like
mypool/xxx/m1
mypool/xxx/m2
etc.
would work, but on the ESX if I mount xxx I just see empty directories for
m1 and m2 as it only works with NFS v3 which does not support automatic
cross mounts.

NetApp has the ability to instantly clone single files and that would also
solve our problem if its somewhere in ZFS road-map (unless the issues we
have above can be resolved)

Thanks,
-- Schachar

On Wed, Apr 21, 2010 at 3:42 PM, David Magda <dmagda at ee.ryerson.ca>
wrote:
> On Wed, April 21, 2010 02:41, Schachar Levin wrote:
> > Hi,
> > We are currently using NetApp file clone option to clone multiple VMs
on
> > our FS.
> >
> > ZFS dedup feature is great storage space wise but when we need to
clone
> > allot of VMs it just takes allot of time.
> >
> > Is there a way (or a planned way) to clone a file without going
through
> > the process of actually copying the blocks, but just duplicating its
meta
> > data like NetApp does?
>
> You mean like the ZFS "clone" command?
>
>      zfs clone [-p] [-o property=value] ... snapshot filesystem|volume
>
> From zfs(1M):
> [...]
>  Clones
>     A clone is a writable volume or file  system  whose  initial
>     contents are the same as another dataset. As with snapshots,
>     creating a clone is nearly instantaneous, and initially con-
>     sumes no additional space.
>
>     Clones can only be created from a snapshot. When a  snapshot
>     is  cloned,  it  creates  an implicit dependency between the
>     parent and child. Even though the clone is created somewhere
>     else  in the dataset hierarchy, the original snapshot cannot
>     be destroyed as long as a clone exists. The origin  property
>     exposes  this  dependency, and the destroy command lists any
>     such dependencies, if they exist.
> [...]
>
> So you''d have something like the following:
>
>   # zfs create /mypool/machine00
>   [ Create VMDK stuff. ]
>   # zfs snapshot /mypool/machine00 at template
>   # for i in 01 02 03 04 05 \
>     do \
>         zfs mypool/machine00 at template mypool/machine${i} \
>     done
>   #
>
> This should work for both regular file systems (shared via NFS) and zvols
> (via iSCSI); at $WORK we''ve used the latter to clone LDoms vdisks.
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20100421/b46ddb38/attachment.html>

If I could mount unlimited NFS mounts on ESX it would be good for me, but there
is a 64 nfs datastore limit.

Right now if I copy a big file it can take several minutes. I already tried
having one VM per zfs file system, and snapshot & clone works great, as long
as I stay below 64 NFS mounts.
Note that VMWARE doesn''t support NFS v4 or else doing something like
mypool/xxx/m1
mypool/xxx/m2
etc.
would work, but on the ESX if I mount xxx I just see empty directories for m1
and m2 as it only works with NFS v3 which does not support automatic cross
mounts.

NetApp has the ability to instantly clone single files and that would also solve
our problem if its somewhere in ZFS road-map (unless the issues we have above
can be resolved)

Thanks,
-- Schachar
-- 
This message posted from opensolaris.org

On Wed, April 21, 2010 09:18, Schachar Levin wrote:
> NetApp has the ability to instantly clone single files and that would also
> solve our problem if its somewhere in ZFS road-map (unless the issues we
> have above can be resolved)
Beyond things like dedupe (and compression), ZFS currently does not have a
per-file capability like that.

ZFS is generally designed to work at either either the pool-level (for
dedupe) or dataset-level (snapshots, cloning, attributes, etc.).

On Apr 20, 2010, at 11:41 PM, Schachar Levin wrote:
> Hi,
> We are currently using NetApp file clone option to clone multiple VMs on
our FS.
In your use case, how do you isolate one VM from another?  
If they are both in the same file system, then they are not isolated. 
If they are not isolated, how do you manage multi-tenancy?
 -- richard

ZFS storage and performance consulting at http://www.RichardElling.com
ZFS training on deduplication, NexentaStor, and NAS performance
Las Vegas, April 29-30, 2010 http://nexenta-vegas.eventbrite.com