Date: Thu, 18 Feb 2010 09:50:32 -0800 From: Matt Reimer <mattjreimer@gmail.com> To: Chris <behrnetworks@gmail.com> Cc: Scot Hetzel <swhetzel@gmail.com>, Pegasus Mc Cleaft <ken@mthelicon.com>, freebsd-current@freebsd.org Subject: Re: Seeing the dreaded "ZFS: i/o error - all block copies unavailable" on 9.0-CURRENT Message-ID: <f383264b1002180950t51ddb421i51b36886e7da6295@mail.gmail.com> In-Reply-To: <64aa03031002180531n5e07e3bcj455ba09bdf5e293e@mail.gmail.com> References: <64aa03031002161803h667734cal4d668b9eb9c0a1a8@mail.gmail.com> <790a9fff1002161842g17de8ecfvf2cfa8c77f03c32@mail.gmail.com> <64aa03031002161849s7b66b9e3l727aafd2afd3d596@mail.gmail.com> <201002171840.41088.ken@mthelicon.com> <790a9fff1002171303u4b40a90cr626ef856efee473b@mail.gmail.com> <64aa03031002171416s128ff196y92a5be5a6abadeb@mail.gmail.com> <64aa03031002171949gdfbd99ci8eab7f29399cc011@mail.gmail.com> <f383264b1002172226s9a76ae6sac55dca22dface77@mail.gmail.com> <64aa03031002180531n5e07e3bcj455ba09bdf5e293e@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, Feb 18, 2010 at 5:31 AM, Chris <behrnetworks@gmail.com> wrote:
> > Can you paste the exact error? Are you getting something like:
> > error 1: lba 32
> > error 1: lba 1
> > When I've seen the above sequence, it was due to a stack overflow (IIRC),
> > with the result that the loader would start a second time and barf out
> these
> > errors.
> > The "large number" in the error might give us a clue as to what's going
> on.
> > If the number is really large, it might be an indication that your BIOS
> > doesn't reliably read past a certain threshold. ZFS writes a sort of
> label
> > at the beginning and end of its drives; perhaps the loader is trying to
> read
> > the label at the end of the disk and is failing (I don't recall whether
> it
> > tries to read both labels).
> > Matt
>
> Matt,
>
> Here's exactly what I'm seeing after POST:
>
> error 1 lba 179552888
> error 1 lba 59243926
> ZFS: i/o error - all block copies unavailable
>
Hmm, both those LBAs fit in 28 bits. With a modern BIOS I wouldn't expect
problems with those LBAs.
It may or may not be related, but I notice that while in this state my
> hard drive activity light stays on.
>
I think you're probably right about this being a BIOS problem.
Does the gptzfsboot 'status' command show the same output when you run it
with and without the F12 workaround? Here's what I see from FreeBSD and from
gptzfsboot:
[root ~]# zpool status
pool: glamdring
state: ONLINE
scrub: none requested
config:
NAME STATE READ WRITE CKSUM
glamdring ONLINE 0 0 0
raidz2 ONLINE 0 0 0
label/glamdring-0 ONLINE 0 0 0
label/glamdring-1 ONLINE 0 0 0
label/glamdring-2 ONLINE 0 0 0
label/glamdring-3 ONLINE 0 0 0
label/glamdring-4 ONLINE 0 0 0
label/glamdring-5 ONLINE 0 0 0
errors: No known data errors
...
/boot.config: -Dh
Default: glamdring:/boot/zfsloader
boot: status pool: glamdring
config:
NAME STATE
glamdring ONLINE
raidz2 ONLINE
label/glamdring-0 ONLINE
label/glamdring-1 ONLINE
label/glamdring-2 ONLINE
label/glamdring-3 ONLINE
label/glamdring-4 ONLINE
label/glamdring-5 ONLINE
Matt
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?f383264b1002180950t51ddb421i51b36886e7da6295>