Re: ZFS and mv utility
- From: Pawel Jakub Dawidek <pjd@xxxxxxxxxxx>
- Date: Wed, 25 Apr 2007 23:24:37 +0200
On Thu, Apr 26, 2007 at 12:56:50AM +0400, Alexey Tarasov wrote:
Hello Pawel.
I have a question about ZFS and mv utility. I have the following pool configuration:
NAME USED AVAIL REFER MOUNTPOINT
storage 62,1G 121G 26K /storage
storage/music 7,95G 121G 7,95G /storage/music
storage/oldhome 47,5G 121G 47,5G /storage/oldhome
I made the following operation: mv /storage/oldhome/lexa/music/* /storage/music and that operation took a very long time. It seems that MV copies files first from
storage/oldhome/lexa/musc to storage/music and then deletes them from old location. But I move files within one storage pool and it should work as fast, as it works on UFS
partition. Does it need to be modified for compatibility with ZFS?
From mv(1) manual page:
As the rename(2) call does not work across file systems, mv uses cp(1)
and rm(1) to accomplish the move. The effect is equivalent to:
rm -f destination_path && \
cp -pRP source_file destination && \
rm -rf source_file
ZFS pool is not a file system. Look at mount(8) output. It would be
possible in theory to move data faster within one pool and this was
discussed on zfs-discuss@xxxxxxxxxxxxxxx, you may want to search the
archives.
--
Pawel Jakub Dawidek http://www.wheel.pl
pjd@xxxxxxxxxxx http://www.FreeBSD.org
FreeBSD committer Am I Evil? Yes, I Am!
Attachment:
pgp3MZNDxIb4R.pgp
Description: PGP signature
- References:
- ZFS and mv utility
- From: Alexey Tarasov
- ZFS and mv utility
- Prev by Date: ZFS and mv utility
- Next by Date: zfs lor, panic
- Previous by thread: ZFS and mv utility
- Next by thread: zfs lor, panic
- Index(es):
Relevant Pages
|
|
