Date: Tue, 20 Jul 2010 22:57:25 -0700 (PDT) From: alan bryan <alanbryan1234@yahoo.com> To: Freddie Cash <fjwcash@gmail.com>, Dan Langille <dan@langille.org> Cc: freebsd-stable <freebsd-stable@freebsd.org> Subject: Re: Problems replacing failing drive in ZFS pool Message-ID: <578438.38753.qm@web50502.mail.re2.yahoo.com> In-Reply-To: <4C4504DF.30602@langille.org>
next in thread | previous in thread | raw e-mail | index | archive | help
=0A=0A--- On Mon, 7/19/10, Dan Langille <dan@langille.org> wrote:=0A=0A> Fr= om: Dan Langille <dan@langille.org>=0A> Subject: Re: Problems replacing fai= ling drive in ZFS pool=0A> To: "Freddie Cash" <fjwcash@gmail.com>=0A> Cc: "= freebsd-stable" <freebsd-stable@freebsd.org>=0A> Date: Monday, July 19, 201= 0, 7:07 PM=0A> On 7/19/2010 12:15 PM, Freddie Cash=0A> wrote:=0A> > On Mon,= Jul 19, 2010 at 8:56 AM, Garrett Moore<garrettmoore@gmail.com>=A0=0A> wrot= e:=0A> >> So you think it's because when I switch from the=0A> old disk to = the new disk,=0A> >> ZFS doesn't realize the disk has changed, and=0A> thin= ks the data is just=0A> >> corrupt now? Even if that happens, shouldn't the= =0A> pool still be available,=0A> >> since it's RAIDZ1 and only one disk ha= s gone=0A> away?=0A> > =0A> > I think it's because you pull the old drive, = boot with=0A> the new drive,=0A> > the controller re-numbers all the device= s (ie da3 is=0A> now da2, da2 is=0A> > now da1, da1 is now da0, da0 is now = da6, etc), and ZFS=0A> thinks that all=0A> > the drives have changed, thus = corrupting the=0A> pool.=A0 I've had this=0A> > happen on our storage serve= rs a couple of times before=0A> I started using=0A> > glabel(8) on all our = drives (dead drive on RAID=0A> controller, remove=0A> > drive, reboot for w= hatever reason, all device nodes=0A> are renumbered,=0A> > everything goes = kablooey).=0A> =0A> Can you explain a bit about how you use glabel(8) in=0A= > conjunction with ZFS?=A0 If I can retrofit this into an=0A> exist ZFS arr= ay to make things easier in the future...=0A> =0A> 8.0-STABLE #0: Fri Mar= =A0 5 00:46:11 EST 2010=0A> =0A> ]# zpool status=0A> =A0 pool: storage=0A> = state: ONLINE=0A> scrub: none requested=0A> config:=0A> =0A> =A0 =A0 =A0 = =A0 NAME=A0 =A0 =A0 =A0=0A> STATE=A0 =A0=A0=A0READ WRITE CKSUM=0A> =A0 =A0 = =A0 =A0 storage=A0=0A> =A0=A0=A0ONLINE=A0 =A0=0A> =A0=A0=A00=A0 =A0=A0=A00= =A0=0A> =A0=A0=A00=0A> =A0 =A0 =A0 =A0 =A0 raidz1=A0 =A0=0A> ONLINE=A0 =A0 = =A0=A0=A00=A0=0A> =A0=A0=A00=A0 =A0=A0=A00=0A> =A0 =A0 =A0 =A0 =A0 =A0 ad8= =A0=0A> =A0=A0=A0ONLINE=A0 =A0=0A> =A0=A0=A00=A0 =A0=A0=A00=A0=0A> =A0=A0= =A00=0A> =A0 =A0 =A0 =A0 =A0 =A0 ad10=A0 =A0=0A> ONLINE=A0 =A0 =A0=A0=A00= =A0=0A> =A0=A0=A00=A0 =A0=A0=A00=0A> =A0 =A0 =A0 =A0 =A0 =A0 ad12=A0 =A0=0A= > ONLINE=A0 =A0 =A0=A0=A00=A0=0A> =A0=A0=A00=A0 =A0=A0=A00=0A> =A0 =A0 =A0 = =A0 =A0 =A0 ad14=A0 =A0=0A> ONLINE=A0 =A0 =A0=A0=A00=A0=0A> =A0=A0=A00=A0 = =A0=A0=A00=0A> =A0 =A0 =A0 =A0 =A0 =A0 ad16=A0 =A0=0A> ONLINE=A0 =A0 =A0=A0= =A00=A0=0A> =A0=A0=A00=A0 =A0=A0=A00=0A> =0A> > Of course, always have good= backups.=A0 ;)=0A> =0A> In my case, this ZFS array is the backup.=A0 ;)=0A= > =0A> But I'm setting up a tape library, real soon now....=0A> =0A> -- Dan= Langille - http://langille.org/=0A>; ______________________________________= _________=0A> freebsd-stable@freebsd.org=0A> mailing list=0A> http://lists.= freebsd.org/mailman/listinfo/freebsd-stable=0A> To unsubscribe, send any ma= il to "freebsd-stable-unsubscribe@freebsd.org"=0A> =0A=0ADan,=0A=0AHere's h= ow to do it after the fact:=0A=0Ahttp://unix.derkeiler.com/Mailing-Lists/Fr= eeBSD/current/2009-07/msg00623.html=0A=0A--Alan Bryan=0A=0A=0A=0A=0A=0A =
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?578438.38753.qm>