Date: Wed, 7 Apr 2010 12:33:34 -0500 From: Dan Nelson <dnelson@allantgroup.com> To: krad <kraduk@googlemail.com> Cc: Freddie Cash <fjwcash@gmail.com>, current@freebsd.org Subject: Re: When will we can use ZFS v24? Message-ID: <20100407173334.GC76941@dan.emsphone.com> In-Reply-To: <j2yd36406631004070900sdd084342kb53331349f3e5c74@mail.gmail.com> References: <m2z78fb9d961004062026kd063f1f3h1b63c3bd7f305289@mail.gmail.com> <r2t11167f521004062058p803f184bs792e94a90b761092@mail.gmail.com> <k2u78fb9d961004062135r67e18238zd16f825dda71dad5@mail.gmail.com> <k2pb269bc571004062138jde60bb10x84da21f328e5a475@mail.gmail.com> <j2yd36406631004070900sdd084342kb53331349f3e5c74@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
In the last episode (Apr 07), krad said: > On 7 April 2010 05:38, Freddie Cash <fjwcash@gmail.com> wrote: > > On Tue, Apr 6, 2010 at 9:35 PM, lhmwzy <lhmwzy@gmail.com> wrote: > > > What's your mean?? > > > > See the archives for the freebsd-fs mailing list. There are two > > separate groups working on getting ZFSv22 added to FreeBSD 9-CURRENT. > > And there's work ongoing to get ZFSv15 added to FreeBSD 8-STABLE. > > > > IOW, just be patient. Good things will come to those who wait. ;) > > If you re hankering for dedup you better have a big box as its a complete > resource hog. From what we have seen on our x4500 filers we need about > 100gig of ram or l2arc ssd to hold the all the metadata for the 34TB of > storage, otherwise the system grinds to a halt. I wish they had an option for "opportunistic" dedup, where if a block is already in ARC it can be a dedupe candidate, but it won't look it up in the DDT if it isn't already in memory. That catches the simple "cp -R dir1 dir2" cases without blowing up the ARC on DDT blocks that 99.9999% of the time don't match. -- Dan Nelson dnelson@allantgroup.com
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20100407173334.GC76941>