zfs discuss - Dec 2005 - zfs consumes 150gb more space than my ext3 server?

Previously I had one server at home running rhel 4 with lvm/ext3 and about 1TB
of storage.

 I built a second server to back that one up with about 1TB and I used zfs for
those file systems.  Once all that was done I used rsync to mv the data over,

ext3
df -h /home/amy
Filesystem            Size  Used Avail Use% Mounted on
/dev/mapper/VolGroup01-LogVol00
                      591G  448G  114G  80% /home/amy

zfs
 zfs list zfs/home/amy
NAME                   USED  AVAIL  REFER  MOUNTPOINT
zfs/home/amy           596G   104G   596G  /export/home/amy

I''m aware I can use compression, not sure how you tell it higher values
or what not, but still from one-to-one should this really be 148gb fatter?

Any thoughts?

Thanks,
This message posted from opensolaris.org

> ext3		448G
> zfs		596G
Educated guess: you''ve set up a 4-disk RAID-Z stripe.
There''s a bug in the way we report space usage for
RAID-Z: it currently includes the parity.  This is
described in a little more detail here:

http://www.opensolaris.org/jive/post!reply.jspa?messageID=15532

I''m guessing that you have a 4-disk RAID-Z because 596G * 3/4 = 447G.

Jeff

>Previously I had one server at home running rhel 4 with lvm/ext3 and about
1TB of storage.
>
> I built a second server to back that one up with about 1TB and I used zfs
for those file systems.
  Once all that was done I used rsync to mv the data
over,
>
>ext3
>df -h /home/amy
>Filesystem            Size  Used Avail Use% Mounted on
>/dev/mapper/VolGroup01-LogVol00
>                      591G  448G  114G  80% /home/amy
>
>zfs
> zfs list zfs/home/amy
>NAME                   USED  AVAIL  REFER  MOUNTPOINT
>zfs/home/amy           596G   104G   596G  /export/home/amy
>
>I''m aware I can use compression, not sure how you tell it higher
values or what not, but still fro
m one-to-one should this really be 148gb fatter?
>
>Any thoughts?

Does rsync handle files with holes and might you have had some?

Casper

Casper.Dik at Sun.COM wrote:
> Does rsync handle files with holes and might you have had some?
AFAIK: No.

The best way to sync two servers with different archs is to use star in 
incremental mode. There is a special flag that makes incremental syncing
more easy.

J?rg

-- 
 EMail:joerg at schily.isdn.cs.tu-berlin.de (home) J?rg Schilling D-13353 Berlin
       js at cs.tu-berlin.de		(uni)  
       schilling at fokus.fraunhofer.de	(work) Blog: http://schily.blogspot.com/
 URL:  http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily

On 12/8/05, Jeff Bonwick <bonwick at zion.eng.sun.com>
wrote:
>
> > ext3          448G
> > zfs           596G
>
> Educated guess: you''ve set up a 4-disk RAID-Z stripe.
> There''s a bug in the way we report space usage for
> RAID-Z: it currently includes the parity.  This is
> described in a little more detail here:
>
> http://www.opensolaris.org/jive/post!reply.jspa?messageID=15532
>
> I''m guessing that you have a 4-disk RAID-Z because 596G * 3/4 =
447G.
>
> Jeff
>
>
There are 4x250gb disk in raidz which gave about 920gb total usable space.

zfs list
NAME                   USED  AVAIL  REFER  MOUNTPOINT
zfs                    715G   208G    16K  /zfs
zfs/home               715G   208G  18.0K  /export/home
zfs/home/amy           596G   104G   596G  /export/home/amy
zfs/home/justin        120G  30.4G   120G  /export/home/justin

# zfs list -o name,compressratio
NAME                  RATIO
zfs                   1.02x
zfs/home              1.02x
zfs/home/amy          1.01x
zfs/home/justin       1.09x

zfs get -o property,value,source all zfs/home/amy
PROPERTY       VALUE                      SOURCE
type           filesystem                 -
creation       Mon Dec  5 20:54 2005      -
used           596G                       -
available      104G                       -
referenced     596G                       -
compressratio  1.01x                      -
mounted        yes                        -
quota          700G                       local
reservation    none                       default
recordsize     128K                       default
mountpoint     /export/home/amy           local
sharenfs       rw                         inherited from zfs/home
checksum       on                         default
compression    on                         local
atime          on                         default
devices        on                         default
exec           on                         default
setuid         on                         default
readonly       off                        default
zoned          off                        default
snapdir        visible                    default
aclmode        groupmask                  default
aclinherit     secure                     default



Can you increase the compressratio ?
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20051208/8e6da387/attachment.html>