Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 20 Jun 2011 21:55:13 -0700
From:      Freddie Cash <fjwcash@gmail.com>
To:        "James L. Lauser" <james@jlauser.net>
Cc:        freebsd-fs@freebsd.org
Subject:   Re: ZFSv28 Dedup question
Message-ID:  <BANLkTinCXPfzqNPS5EGEw1b46mM9FaVD5g@mail.gmail.com>
In-Reply-To: <BANLkTikZZg3W4zRED-Nj88y6kco_1qBehz2XJ2wq=Wj4Zj3Jsw@mail.gmail.com>
References:  <BANLkTikZZg3W4zRED-Nj88y6kco_1qBehz2XJ2wq=Wj4Zj3Jsw@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Jun 20, 2011 at 7:46 PM, James L. Lauser <james@jlauser.net> wrote:

> I have a (hopefully) simple question about ZFSv28 deduplication that is not
> very clearly answered in any of the documentation I've found.
>
> If I enable dedupe on a dataset, will newly written data dedupe against old
> data that was written before dedupe was enabled?  If not, is it worth
> forcing the data to be rewritten so that it makes it into the appropriate
> tables?
>
> No.  Dedupe only applies to newly written data.  The DDT (dedupe table)
only contains entries for data written after the "dedup" property is set on
a filesystem.

The only way to dedupe existing data is to copy/delete it, or "zfs send|zfs
recv" it, or otherwise re-write it.
-- 
Freddie Cash
fjwcash@gmail.com



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?BANLkTinCXPfzqNPS5EGEw1b46mM9FaVD5g>